Anti–Glutamic Acid Decarboxylase Antibodies in the Serum and Cerebrospinal Fluid of Patients With Stiff-Person Syndrome: Title and subTitle BreakCorrelation With Clinical Severity FREE

Stiff-person syndrome (SPS) is an immune-mediated central nervous system disorder characterized by rigidity of the axial and proximal limb muscles, intermittent superimposed spasms, and heightened sensitivity to external stimuli.^{1 - 3} Continuous cocontraction of agonist and antagonist muscles caused by involuntary firing of motor units at rest are the clinical and electrophysiologic hallmarks of the disease.^{1 - 4}

Stiff-person syndrome is an autoimmune disorder that responds to immunotherapies^{5 - 6} and is frequently associated with other autoimmune diseases, autoantibodies, and certain HLA haplotypes.^{1 - 3} A characteristic autoimmune marker for SPS seen in more than 80% of the patients is the presence of antibodies against glutamic acid decarboxylase (GAD65), the rate-limiting enzyme for the synthesis of γ-aminobutyric acid (GABA), the brain's main inhibitory neurotransmitter.^{1 - 3} These antibodies are produced intrathecally and may impair the synthesis of GABA, resulting in low GABA levels in the brain and cerebrospinal fluid (CSF).^{2 - 3} To determine the clinical relevance of anti-GAD antibodies in patients with SPS, we correlated their intrathecal production and serum titers with the severity of symptoms, using validated clinical scales.

We examined serum and CSF from 16 patients with SPS, 6 men and 10 women, aged 37 to 62 years (mean, 52 years), studied under institutional review board–approved protocols at the Neuromuscular Diseases Section, National Institutes of Health, Bethesda, Md, from 1997 to 2003. All patients had elevated titers of serum anti-GAD antibodies and fulfilled the established diagnostic criteria for SPS.^{1 ,3} The serum and CSF were examined for total immunoglobulins, immunofixation electrophoresis, IgG index, and intrathecal synthesis of IgG-specific anti-GAD antibodies. The integrity of the blood-brain barrier was assessed by the ratio of serum albumin to CSF albumin (normal, >130). The CSF IgG index was calculated according to the following formula:

(CSF IgG/Serum IgG)/(CSF Albumin/Serum Albumin),

in which CSF IgG and CSF albumin are measured in nanograms per milliliter and serum IgG and serum albumin are measured in milligrams per milliliter.

The CSF IgG index is considered normal when it is below 0.67.^{2 ,7} For detection and quantification of anti-GAD antibodies in serum and CSF, we used a radioimmunoassay kit (Kronus, Boise, Idaho). The results are reported in manufacturer's units per milliliter (normal, <1 U/mL). For the calculation of IgG GAD–specific index, the following formula was used in paired serum and CSF specimens:

(CSF IgG Anti-GAD Titer/Serum IgG Anti-GAD Titer)/(CSF Albumin/Serum Albumin),

in which the CSF IgG and serum IgG anti-GAD titers are measured in units per milliliter, CSF albumin is measured in nanograms per milliliter, and serum albumin is measured in milligrams per milliliter.

The final concentration of antibodies was obtained after normalization of the different dilutions to the same standard curve. Antibody titers were blindly analyzed and compared with the stiffness index and heightened sensitivity scores, determined for each patient at the time of antibody draw.

The following scales that measure disease severity were used, as previously described.^{1 - 2 ,6} For heightened sensitivity, the following were rated as present or absent: noise-induced stiffness and cramps, visual stimuli–induced stiffness and cramps, somatosensory-induced stiffness and cramps, voluntary activity–induced spasms, emotional upset and "stress"-induced spasms, awakenings due to nocturnal spasms, and untriggered cramps and spasms. The presence of each item added 1 point (maximum score, 7). For distribution of stiffness and cramps, the presence of each of the following was given 1 point: stiffness in the face, stiffness in the arms, stiffness in the upper trunk, stiffness in the abdomen, stiffness in the lower trunk, and stiffness in the legs (maximum score, 6).

The duration of disease ranged from 5 to 30 years (mean, 11 years); the time from onset of symptoms to diagnosis ranged from 1 to 13 years (mean, 4 years). The most prevalent initial symptom was lower back muscle stiffness (8 patients), followed by proximal leg stiffness (7 patients). Single-limb stiffness was the initial symptom in 1 patient, and stiffness confined to respiratory and abdominal muscles in another. The median score on the stiffness index was 3.3 (range, 1-6) (Figure 1). The median score for heightened sensitivity was 3.8 (range, 2-7). At the time of GAD antibody determination, all studied patients were symptomatic or incapacitated, despite receiving GABA-enhancing drugs such as diazepam, clonazepam, or baclofen. None of the study patients were taking immunomodulating therapies, but 1 patient had received intravenous immunoglobulin 4 months before determination of GAD antibodies. Seven patients also had insulin-dependent diabetes mellitus, 6 had immune-mediated thyroiditis, and 2 had pernicious anemia. Five patients had strongly positive antinuclear antibodies, 3 had thyroid microsomal antibodies, and 2 others had anti-islet or anti-Jo antibodies.

Neuropsychiatric symptoms were ubiquitous, with anxiety being most prevalent in 9 patients and depressed mood in 3. Five patients had seizures.

No consistent correlation was found between the serum anti-GAD antibody titers and disease severity; the titers were high in some patients with mild disease and low in some others with severe disease (Figure 1).

The mean anti-GAD antibody titer in the CSF was 181 U/mL (range, 30-400 U/mL) and the mean serum titer was 51 500 U/mL (range, 24 000-200 000) (normal, <1 U/mL). The mean ratio of the CSF to serum anti-GAD antibody titer was 0.02 (range, 0.001-0.166), compared with the mean ratio of 0.002 for the CSF to serum IgG (range, 0.001-0.004), reflecting a 10-fold increased production of GAD-specific IgG antibodies in the CSF.

The CSF to serum albumin ratio was normal in all patients, indicating intact blood-brain barrier. No correlation was found between the CSF anti-GAD antibody titers and their intrathecal production with severity of clinical symptoms (Figure 2). Although 6 patients with high intrathecal anti-GAD antibody production had clinically severe disease, others with equally severe disease had low GAD-specific CSF IgG index.

The level of anti-GAD antibodies in serum or CSF did not correlate with the presence of other autoantibodies or the patients' subsequent response to intravenous immune globulin.⁶ Preliminary results in 4 patients who had anti-GAD antibodies measured serially during a 2-year period demonstrated that titers did not correlate with the clinical fluctuations of their disease (data not shown).

High anti-GAD antibodies appear to be specific markers for SPS; only rare patients with cerebellar degeneration and diabetes have high antibody titers.⁸ It has been proposed that these antibodies may be pathogenic because (1) GAD activity and GABA synthesis are directly inhibited in vitro by the anti–GAD-specific IgG from patients with SPS⁹ and (2) anti-GAD antibodies are produced intrathecally and may suppress GABA level, as supported by the finding of reduced GABA in the brain and CSF.^{2 - 3} Because GABA is the brain's predominant inhibitory neurotransmitter, reduction of GABA could explain the patient's symptoms of muscle hyperactivity, as well as the motor cortex hyperexcitability noted by the electrophysiologic studies.⁴ Our finding that the anti-GAD antibodies, which recognize a cytoplasmic antigen, do not correlate with the clinical symptoms casts doubt on their direct pathogenic role in affecting GABA level. Whether GABA levels correlate with disease severity remains unclear.

Glutamic acid decarboxylase is a cytoplasmic antigen, not easily recognized by the immune system, indicating that another surface protein associated with GAD or GABA may be the target of the autoimmune attack. Although GAD antibodies may cross-react with GABA binding sites on the neuronal surface and GAD peptide fragments may be expressed at the cell surface during exocytosis of GABA, there is no sufficient evidence to prove their pathogenic role in SPS. Alternatively, cytotoxic T cells may be the main pathogenic effectors if cytoplasmic GAD peptides, bound to major histocompatibility complex I molecules, are recognized by the CD8⁺ T cells, as proposed for paraneoplastic neuronal antibodies.¹⁰

Until the specific surface antigen responsible for the initiation of self-sensitization is demonstrated, the anti-GAD antibodies in SPS should be considered only as excellent disease markers. Our present findings indicate that there is no practical value in observing anti-GAD antibodies serially during the various stages of the disease or to monitor response to therapies. The observation is consistent with the significance of autoantibody levels in other autoimmune disorders such as myasthenia gravis or Lambert-Eaton myasthenic syndrome.

Corresponding author and reprints: Marinos C. Dalakas, MD, Neuromuscular Diseases Section, NINDS, NIH, Bldg 10, Room 4N248, 10 Center Dr, MSC 1382, Bethesda, MD 20892-1382 (e-mail: dalakasm@ninds.nih.gov).

Accepted for publication November 24, 2003.

Author contributions: Study concept and design (Drs Rakocevic and Dalakas); acquisition of data (Drs Rakocevic and Raju); analysis and interpretation of data (Drs Rakocevic, Raju, and Dalakas); drafting of the manuscript (Dr Rakocevic); critical revision of the manuscript for important intellectual content (Drs Rakocevic, Raju, and Dalakas); statistical expertise (Dr Rakocevic); obtained funding (Dr Dalakas); administrative, technical, and material support (Drs Rakocevic and Dalakas); study supervision (Drs Raju and Dalakas).

This study was funded through the intramural program of the National Institute of Neurological Disroders and Stroke (Dr Dalakas).