Stiff-person syndrome

INTRODUCTION — Stiff-person syndrome (SPS, formerly called stiff-man syndrome) is an uncommon disorder characterized by progressive muscle stiffness, rigidity, and spasm involving the axial muscles, resulting in severely impaired ambulation [1,2]. It is caused by increased muscle activity due to decreased inhibition of the central nervous system (CNS) that results from the blockade of glutamic acid decarboxylase (GAD), an enzyme critical for maintaining inhibitory pathways. The subsequent decline in the levels of gamma amino butyric acid (GABA) in the CNS causes a loss of neural inhibition [3]. SPS is often associated with type 1 diabetes mellitus (T1DM), which may be a reflection of shared pathogenetic features, as well as other autoimmune disorders. It may rarely occur as a paraneoplastic syndrome.

Classic SPS and its major variants, partial SPS and paraneoplastic SPS, are reviewed here. Progressive encephalomyelitis with rigidity and myoclonus (PERM) and other paraneoplastic disorders of the nervous system are discussed in detail separately. (See "Overview of paraneoplastic syndromes of the nervous system" and "Paraneoplastic syndromes affecting spinal cord, peripheral nerve, and muscle" and "Autoimmune (including paraneoplastic) encephalitis: Clinical features and diagnosis".)

PATHOGENESIS — Several observations suggest an autoimmune component to the pathogenesis of stiff-person syndrome (SPS, formerly called stiff-man syndrome), including an association with type 1 diabetes mellitus (T1DM) and other autoimmune disorders, and the presence of autoantibodies targeting proteins associated with gamma amino butyric acid (GABA)-related neural pathways. It is thought that impairment of GABAergic pathways by autoantibodies and a reduction of brain GABA can lead to the clinical manifestations of stiffness, spasms, and phobias [4]. (See 'Autoimmunity' below and 'Type 1 diabetes' below.)

An autoimmune process was originally postulated to underlie SPS, as the disorder often occurs in conjunction with any one of a variety of autoimmune diseases, particularly T1DM; other associated conditions include thyroiditis, vitiligo, and pernicious anemia [1,5-8]. An association between anti-glutamic acid decarboxylase (GAD) antibodies and SPS has also been noted [8,9]. These antibodies were found to target GABAergic neurons and their nerve terminals. The dominant antigen recognized by these antibodies is the GABA-synthesizing enzyme GAD [10].

Autoimmunity — The pathogenic role of the anti-GAD antibodies in SPS remains to be defined. GAD is an intracellular enzyme and is not readily accessible to binding by antibodies, although transfer experiments in mice, using serum obtained from affected patients, resulted in recipients becoming stiff. However, in SPS the titer of anti-GAD antibodies does not always correlate with disease activity, and about 30 percent of patients are antibody-negative.

Other antibodies that have been identified in patients with SPS include:

●Antibodies directed against the 14 kDa postsynaptic GABA_A receptor-associated protein have been detected in almost 70 percent of patients with non-paraneoplastic SPS [11]. This protein interacts with gephyrin, another protein that may be antigenic in some patients with paraneoplastic forms of SPS, enabling the assembly of GABA_A-receptor into the plasma membrane.

●Antibodies directed against the 128 kDa protein amphiphysin, an intracellular protein that promotes cleavage of intracellular vesicles, have been documented in patients with paraneoplastic SPS particularly associated with breast cancer [12].

●Oligoclonal bands (OCB) targeting different GAD epitopes are frequently encountered in the cerebrospinal fluid (CSF) of patients with SPS. These OCBs have a higher frequency and intensity in the CSF, estimated to be 10-fold higher in the central nervous system (CNS) than in peripheral tissues [13].

Whether T cells have an important role in the pathogenesis of SPS remains uncertain. In one study, several GAD-specific T-cell clones that were isolated from the CSF of patients with SPS and intrathecal anti-GAD immunoglobulin G (IgG) were shown to produce Th2 cytokines that could support an antigen-driven collaboration between T and B cells, thus leading to sustained intrathecal production of oligoclonal anti-GAD IgG antibodies [14]. Further support for T-cell involvement was found in an experimental model in which mice possessing a monoclonal GAD65-specific CD4(+) T-cell population, even in the absence of B cells, developed a severe encephalomyelitis-like response [10].

Type 1 diabetes — T1DM is observed in about 30 percent of patients with SPS, but only 0.01 percent of patients with T1DM have SPS. Anti-GAD65 antibodies are also found in T1DM, being present in up to 80 percent of patients and in first-degree relatives, where they serve as a marker for the disease and as a risk predictor. They are often found in low titers in the serum of patients before the onset of diabetes. (See "Pathogenesis of type 1 diabetes mellitus" and "Type 1 diabetes mellitus: Disease prediction and screening".)

There is a significant difference between T1DM and SPS in the epitope specificity of the anti-GAD antibodies in these two conditions [15,16]. In T1DM the antibodies are thought to recognize conformational epitopes, while the antibodies in SPS mostly recognize linear and denatured epitopes especially in the amino terminal region of the GAD65 molecule [17].

Since SPS appears to be an autoimmune disorder with links to type 1 diabetes, it is not surprising that the inheritance of certain human leukocyte antigen (HLA) alleles, particularly those implicated in type 1 diabetes, may be associated with increased risk of developing SPS. In one study, significantly more patients with SPS (13 of 18) than controls (18 of 48) carried the DQB1*0201 allele (72 versus 38 percent) [18]. In addition, other DQB1 alleles rarely found in type I diabetes (such as *0602 and sequence-related *06 alleles) were found to have similar prevalence in patients with SPS and controls, with a lower proportion of such patients also having T1DM, suggesting that unlike DQB1*0201, the *06 alleles may be associated with a reduced prevalence of diabetes in patients with SPS. (See "Pathogenesis of type 1 diabetes mellitus".)

CLINICAL MANIFESTATIONS

Epidemiology — Stiff-person syndrome (SPS, formerly called stiff-man syndrome) has an estimated prevalence of one to two cases per million, with an incidence of one case per million per year [4]. Most patients present between the ages of 20 and 50, and women are affected two to three times more often than men. It occurs extremely rarely in childhood [19].

Clinical subtypes of SPS — Patients with SPS have been subdivided into three categories, based upon the presence or absence of specific antibodies and other diseases:

●Classic SPS – These patients present with truncal stiffness, generalized rigidity, and frequent muscle spasms, resulting in an awkward, wide-based gait. The majority of patients with SPS (70 to 80 percent) present in this fashion.

●Partial SPS – The most common form of partial SPS is stiff-limb syndrome, in which patients have marked difficulty with ambulation due to stiffness and lack of mobility affecting one limb, generally a leg. Other localized forms may include abdominal or chest wall muscle involvement, causing severe painful spasms or ocular symptoms [20]. This presentation is noted in about 10 to 15 percent of patients.

●Paraneoplastic SPS variant – These patients are clinically indistinguishable from patients with classic SPS. Though they are usually glutamic acid decarboxylase (GAD) antibody-negative, antibodies targeting specific proteins such as amphiphysin have been detected. This is an extremely rare form, but this diagnosis should always be considered in patients who are GAD antibody-negative. It is seen in less than 2 percent of patients with SPS. (See 'Paraneoplastic SPS' below.)

Progressive encephalomyelitis with rigidity and myoclonus (PERM), a rare condition that can co-occur with SPS, is seen in less than 0.1 percent of patients. (See 'Progressive encephalomyelitis with rigidity and myoclonus' below.)

Classic SPS — In its classical form, SPS is characterized by excessive muscle rigidity of the lumbar, trunk, and proximal limb muscles that is caused by sustained muscular contractions occurring in agonist and antagonist muscles. The clinical hallmark of SPS is an extreme degree of muscle stiffness and rigidity. The persistent nature of these symptoms may lead to fixed spinal deformities such as a pronounced lumbar or cervical lordosis; this is thought to be due to the simultaneous contraction of opposing paraspinal muscle groups.

The onset of stiffness and rigidity in the axial muscles, either lumbar or cervical, is insidious and generally progresses slowly over time to involve proximal limb muscles. Patients walk with a wide-based and unsteady stance ("Frankenstein's gait") with a tendency to fall in a fashion similar to a log tumbling down. They are at high risk for fracture [21]. As patients lose their ability to walk safely, they may develop a heightened sense of fear and anxiety and agoraphobia. As truncal flexibility is lost, ambulation slows considerably; the patient may eventually require assistive devices to walk. Activities of daily living, such as getting into or out of bed, arising from a chair, or dressing, are severely limited; some patients become bedridden.

Superimposed episodic muscle spasm precipitated by sudden movement, noise, or emotional upset is a sensitive and specific feature of SPS, known as the startle reflex [22]. Spasms usually begin in the axial muscles and may spread to the extremities. They are painful and can generate sufficient force to fracture bone. The muscle spasm is generally easily visualized; the affected area has a tight, rock-hard appearance and a unique, board-like feel [23]. Palpation may provoke even more intense spasm.

The head retraction reflex (HRR) is a vestigial withdrawal reflex of the face and is suppressed in healthy subjects. In one study, 17 of 28 SPS patients demonstrated a positive HRR, suggesting that the presence of this finding in the setting of a patient with unexplained stiffness and spasms might help to avoid the otherwise frequent misdiagnosis of a psychogenic motor disturbance [24].

Paroxysmal autonomic dysfunction, characterized by transient hyperpyrexia, diaphoresis, tachypnea, tachycardia, pupillary dilatation, and arterial hypertension, has been described and may result in sudden death [25]. Some of these deaths have been attributed to respiratory problems, including sudden apnea with cyanosis, tachypnea, and respiratory arrest; these difficulties may result from diaphragmatic spasm, impaired respiratory function, and severe respiratory muscle rigidity [26,27]. Esophageal dysmotility and swallowing difficulties may lead to aspiration [28].

Other than the abnormal gait and hyperreflexia, the motor and sensory nerve examinations are usually normal.

Though cognition is normal, specific phobia is a frequent non-motor symptom of SPS. In one study, 19 of 44 patients (44.2 percent) developed task-specific phobia, that is, fear and avoidance of situations difficult to master owing to the motor symptoms of SPS (such as crossing streets) [29]. Anxiousness may further exacerbate muscle stiffness or may trigger spasms; and anxiety, panic attacks, and avoidance behavior may lead to a misdiagnosis of a phobic disorder [30,31]. The progressive clinical deterioration contributes to the development of depression in many patients [20,32,33].

Dyspnea is often present in patients with SPS. A prospective study of 17 patients with SPS observed that dyspnea was common and occurs at rest, with exertion, and disturbs sleep [34]. Pulmonary function studies identified a restrictive lung physiology and correlation between pulmonary function variables. The findings support the hypothesis that thoracic cage constriction by rigidity and/or spasm of the muscles of the trunk causes or contributes to the sensation of dyspnea.

Partial SPS — Partial SPS is characterized by the stiff-limb syndrome, in which one extremity, usually a lower limb (termed stiff-leg syndrome [SLS]), is affected, and the trunk is generally spared [35]. The SLS is characterized by focal rigidity of one leg and differs from SPS in various ways [36,37]: contractions and rigidity are limited to the affected extremity, it is rarely associated with a paraneoplastic etiology, and anti-GAD antibodies are positive in only 15 percent of cases of SLS. (See 'Laboratory findings' below.)

Despite the limited involvement, SLS can be associated with a significant functional decline because of the major difficulty faced by the patient who is trying to weight-bear on one completely stiff, non-bending leg and a normal contralateral limb.

Medullary and cerebral magnetic resonance imaging (MRI) do not usually show any abnormalities [38], although rare cases of an associated transverse myelitis have been described.

Paraneoplastic SPS — Patients with the paraneoplastic form of SPS may present clinically with classical features of SPS; however, unlike the majority of patients with classic SPS, those with the paraneoplastic form are usually anti-GAD antibody-negative and often present with concurrent features of an underlying neoplasm, such as anorexia, weight loss, or a suspicious mass. Less than 1 to 2 percent of patients with SPS have paraneoplastic SPS. (See "Paraneoplastic syndromes affecting spinal cord, peripheral nerve, and muscle", section on 'Stiff-person syndrome'.)

Malignancies associated with SPS include breast and lung cancer and Hodgkin lymphoma. Some patients with SPS and solid tumors have antibodies that target amphiphysin, a 128-kd protein concentrated in nerve terminals, where a pool of both GAD and amphiphysin is associated with the cytoplasmic surface of synaptic vesicles [12,39,40]. Passive transfer of immunoglobulin G (IgG) from an affected patient with anti-amphiphysin antibodies has reproduced clinical features of SPS in rats, further supporting a role for autoantibodies in the pathogenesis of this disorder [41]. Ultrastructural analysis of spinal cord presynaptic tissue performed after in vivo intrathecal passive transfer of affinity-purified human anti-amphiphysin autoantibodies in rats demonstrated an autoantibody-induced structural disorganization in GABAergic (gamma amino butyric acid) synapses, with profound changes in presynaptic vesicle pools, activation of alternative endocytic pathways, and potentially compensatory rearrangement of proteins involved in endocytosis [42].

Malignant disease was less common in patients with anti-GAD antibodies in a series of 99 patients with classic SPS or an SPS variant [35]. Among 79 GAD65 antibody-positive patients, three (4 percent) had carcinoma (thyroid, renal cell, and colon); all of these patients had classic SPS, and cancer was found in two of them after the onset of neurological symptoms. Among 20 GAD65 antibody-negative patients, five (25 percent) had cancer. Three of these patients had amphiphysin antibody-positivity and breast carcinoma, and two (one with classic SPS and one with partial SPS) had non-Hodgkin lymphoma.

In one retrospective study, patients with high titers of GAD antibodies who did not demonstrate the classical clinical features of SPS were at greater risk for having a paraneoplastic syndrome. The cancer risk increased with age, male sex, and presence of concomitant antibodies against neuronal cell-surface antigens [43].

Remission of neurologic symptoms has been noted following tumor excision and treatment with glucocorticoids [40,44,45]. This observation suggests that neurologic symptoms are due to functional rather than structural changes in the central nervous system (CNS).

Progressive encephalomyelitis with rigidity and myoclonus — PERM can also resemble SPS; transient oculomotor disturbances are noted in these patients [46]. Continuous electromyographic (EMG) activity at rest is a feature of PERM [38]. It has been proposed that PERM and SPS may be part of a spectrum of encephalomyelopathies displaying autoimmunity against GABAergic neurons, since some patients with PERM are anti-GAD antibody-positive. Overlapping features of PERM and SPS have also been noted in anti-GAD-negative patients [47,48]. One such patient had anti-Ri (antineuronal nuclear autoantibody type 2, or ANNA-2) antibodies present in her serum [47], and the other anti-glycine receptor antibodies [48]. (See "Paraneoplastic syndromes affecting spinal cord, peripheral nerve, and muscle", section on 'Stiff-person syndrome' and "Autoimmune (including paraneoplastic) encephalitis: Clinical features and diagnosis", section on 'Anti-GlyR encephalopathy'.)

A retrospective study of 121 patients with SPS spectrum disorders detected glycine receptor (GlyR) antibodies in about 20 percent of patients. About half of these patients displayed all or partial elements of PERM, including features of brainstem dysfunction, myoclonus, upper or lower motor neuron symptoms, sensory deficits, sphincter or autonomic dysfunction, seizures, and cognitive deficits [49].

Laboratory findings — SPS is not associated with any abnormalities in most routine laboratory studies; muscle enzymes are sometimes mildly elevated. A variety of autoantibodies may be present, particularly anti-GAD antibodies, which are present in a majority of patients. Cerebral spinal fluid studies (CSF) frequently exhibit oligoclonal banding.

●General laboratory studies – In the absence of another diagnosis, patients with all forms of SPS have normal complete blood counts and markers of inflammation such as serum C-reactive protein (CRP) and the erythrocyte sedimentation rate (ESR). The serum creatine kinase and aldolase levels are normal or only slightly elevated; values that are greater than twice the upper limit of normal suggest the possibility of another diagnosis. Other clinical chemistry values are typically normal. Given the propensity for type 1 diabetes mellitus (T1DM) in SPS, some patients may have an elevated glycosylated hemoglobin level.

●Anti-GAD antibodies – Antibodies reacting to GAD, mainly the 65 kDa isoform, are found in high titers (>1000 units/mL) in 60 to 80 percent of patients with classic SPS and are frequently present in the CSF, resulting in oligoclonal banding. Patients with SPS have anti-GAD antibodies recognizing discontinuous segments of the middle and C-terminal part of GAD65, which represents epitopes influenced by molecular conformation [9]. Anti-GAD antibodies are positive in only 15 percent of cases of partial SPS.

In one study, anti-GAD65 antibodies from patients with SPS were shown to be directed against a linear epitope in the N-terminal segment, especially within the first 100 amino acids of the protein, which was not recognized by the antibodies in the serum of patients with T1DM [50]. This segment of GAD is exposed during synaptic transmission, but whether these antibodies are pathogenic is unknown [51].

Oligoclonal bands (OCBs) are frequently encountered in the CSF of patients with SPS; these OCBs have a higher frequency and intensity in the CSF than in the serum and include various subsets of antibodies directed against different GAD epitopes [50]. The production of these oligoclonal antibodies has been estimated to be 10-fold higher in the CNS than in peripheral tissues [52].

●Other autoantibodies – An autoantibody observed less often than anti-GAD may target the alpha-1-subunit of the GlyR in some patients with SPS variants [49,53] (see 'Progressive encephalomyelitis with rigidity and myoclonus' above). A variety of other autoantibodies have also been identified in the sera of patients with SPS, including antithyroid, -thyroid peroxidase, -thyroglobulin, and anti-gastric parietal antibodies. An association between thymoma, myasthenia gravis, and SPS has been noted [54,55]. Following thymectomy and three courses of intravenous immune globulin (IVIG), symptoms of SPS resolved in one patient with this combination of conditions [55].

Electromyographic studies — EMG studies reveal continuous motor-unit activity which is typically decreased or abolished by intravenous diazepam, sleep, and local or general anesthesia [7]. There are no distinguishing electrophysiologic features between GAD antibody-positive and -negative patients [56].

In one detailed study involving three patients with SPS that used simultaneous video-electroencephalographic-surface EMG recordings in addition to routine EMG, the following features were described [57]:

●Continuous muscle activity that varied with waking and sleep states, passive and active movements, and medications

●Superimposed intermittent, generalized contractions while awake, which continued into drowsiness and interfered with the onset of sleep

●Abnormal co-contractions of antagonistic muscles

●A unique motor response, consistent with a stimulus-induced truncal myoclonus, which was observed in all patients

The stimulus-induced truncal myoclonus was not observed in normal subjects or in patients with other disorders causing increased muscle tone. It consisted of a sequence of one to three synchronous myoclonic bursts, lasting 60 to 70 milliseconds after median nerve stimulation, which was followed by a tonic decrescendo activity. The recruitment order of muscles along the neuraxis in this response suggested that it was generated in the spinal cord and was conveyed peripherally via propriospinal tracts. This motor response may serve not only as a diagnostic tool but also may explain some aspects of the pathophysiology of both spasms and stiffness in SPS.

In patients with SPS, unexpected noises often trigger violent muscle jerks resulting in falls. The pattern of EMG-onset latency in these patients suggests that the audiogenic muscle jerks in SPS originate in the acoustic startle response [58].

In addition, widespread altered inhibitory mechanisms have been documented in patients with SPS, with evidence supporting a predominant premotor and supraspinal mechanisms driving the increased-motor activity [3].

Neuropathology — Pathological changes show little or no correlation with clinical manifestations, but few detailed neuropathologic studies of patients with SPS have been reported [25]. Some autopsy studies of patients with SPS show no inflammatory infiltrates or other important pathological changes. Other studies have found chromatolysis and vacuolization of anterior horn cells, more commonly in the caudal levels of the spinal cord and a loss of alpha-motor, gamma-motor neurons and spinal interneurons with gliosis [59].

Neuroimaging — No specific neuroradiologic findings have been noted in SPS.

DIAGNOSIS — Establishing the diagnosis of stiff-person syndrome (SPS, formerly called stiff-man syndrome) requires a high index of suspicion, and evaluation by an expert in neuromuscular or neuroimmune disorders is usually required. The presence of the following features are generally considered necessary for making the diagnosis, although there are no formally accepted criteria [3,7]:

●Stiffness in the axial and limb muscles resulting in impairment of ambulation

●Presence of superimposed episodic spasms that are precipitated by sudden movement, noise, or emotional upset

●A positive therapeutic response to oral diazepam or findings of continuous motor-unit activity on electromyography (EMG) that are abolished by intravenous diazepam

●Absence of other neurologic disorders that may explain the clinical features

We perform the following diagnostic evaluation, in addition to the medical history and physical examination, with particular attention to the neurologic symptoms and findings in general and features characteristic of the disease (see 'Clinical subtypes of SPS' above):

●Anti-glutamic acid decarboxylase (GAD) antibody testing, which is positive in about two-thirds of patients. Those patients with markedly elevated titers are more likely to have SPS.

●Therapeutic trial of diazepam (beginning at a dose of 5 mg four times daily [taken orally] and slowly titrated upwards to the maximum dose required to control symptoms without causing excess sedation), if the response to benzodiazepines has not already been adequately assessed. Doses achieved may be as high as 120 mg per day. (See 'Initial/symptomatic therapy' below.)

●EMG testing, including evaluation of the response to the administration of oral diazepam. Patients with SPS will describe definite reduction in their pain and stiffness allowing for more enhanced physical function.

●Routine laboratory tests should include a complete blood count, chemistry profile, thyroid function, glycosylate hemoglobin (hemoglobin A1C) level, creatine kinase, and C-reactive protein (CRP) to help exclude other conditions that may mimic some of the stiffness features described in patients with SPS. (See 'Differential diagnosis' below.)

When present, anti-GAD antibodies support the diagnosis; however, these antibodies may be absent in up to one-third of patients. In addition, in patients with characteristic clinical findings, anti-GAD antibodies, and a positive response to diazepam, the diagnosis can be made without an EMG, although such electrodiagnostic studies and a clinical response to diazepam can provide additional support for the diagnosis. Once the diagnosis has been established, additional testing is performed to identify or exclude associated conditions. (See 'Post-diagnostic evaluation' below.)

Muscle hardening to a board-like sensation is probably the most specific clinical observation that establishes the diagnosis. In addition, most clinicians rely upon the response to oral diazepam. Patients with SPS will generally note improvement with titrated doses of a benzodiazepine (eg, diazepam [5 to 20 mg taken orally two to four times per day]). Many patients will require high doses to note improvement. (See 'Initial/symptomatic therapy' below.)

In patients with partial SPS, the symptoms are limited to the affected area, typically a single extremity, but otherwise the diagnostic criteria are the same as for most SPS patients. The diagnosis of partial SPS may be more challenging. Focal spinal cord lesions that may impair neurologic function in a limb or intraarticular mechanical derangement (eg, hip, knee, or ankle) that restricts motion across a major joint should be excluded in such patients.

The diagnosis of paraneoplastic SPS can be challenging as well, and a high index of suspicion is required. Patients may present with features of occult malignancy and laboratory tests may provide additional clues. Since GAD antibody positivity is typically not seen in patients with paraneoplastic SPS, any patient who is GAD antibody-negative should be carefully evaluated for occult malignancy.

DIFFERENTIAL DIAGNOSIS — The pain and stiffness described in stiff-person syndrome (SPS, formerly called stiff-man syndrome) may mimic other conditions, including musculoskeletal and neurologic disorders. These include:

●Ankylosing spondylitis – Patients with axial spondyloarthritis, including ankylosing spondylitis (AS), may describe difficulty arising in the morning and have trouble walking because of truncal stiffness that may resemble the stiffness in SPS. AS may be distinguished from SPS by features of the medical history and physical examination characteristic of AS, but not SPS, including improvement following hot showers, physical activity, or the use of nonsteroidal antiinflammatory drugs (NSAIDs) or tumor necrosis factor (TNF)-inhibitors. In addition, lab abnormalities including a raised erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP) and imaging features of sacroiliitis or spondylitis may be observed in AS but are not features of SPS. (See "Diagnosis and differential diagnosis of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults".)

●Parkinson disease – Parkinson disease (PD) may present with truncal stiffness and gait dysfunction that is similar to SPS. A careful neurologic exam should identify cogwheel rigidity in PD. The absence of the startle reflex and the lack of improvement following a challenge with a benzodiazepine further support the diagnosis of PD rather than SPS. Changes on electromyography (EMG) that are not seen in PD can help to confirm the diagnosis of SPS. (See "Diagnosis and differential diagnosis of Parkinson disease".)

●Tetanus – The spasms of the trunk and limbs seen in SPS can simulate the clinical findings in tetanus. The absence of trismus or facial spasms and the rapid response to benzodiazepines (eg, diazepam) that are characteristic of SPS help to distinguish SPS from true tetanospasms [60]. (See "Tetanus".)

●Axial dystonia – These patients may present with dystonia that can resemble partial SPS. The conditions can often be challenging to separate. To help distinguish the two conditions, the presence of glutamic acid decarboxylase (GAD) antibody positivity can be helpful. In addition, the response to benzodiazepines is often more striking in partial SPS than in patients with dystonia [60]. (See "Etiology, clinical features, and diagnostic evaluation of dystonia".)

●Focal limb or joint disorders – Focal spinal cord lesions that may impair neurologic function in a limb and intraarticular mechanical derangement (eg, hip, knee, or ankle) that restricts motion across a major joint may each mimic partial SPS affecting only the extremities. These conditions can be distinguished from SPS by careful history and examination and, where necessary, by neuroimaging and electrodiagnostic studies including EMG. (See "Disorders affecting the spinal cord" and "Approach to the adult with unspecified knee pain".)

●Other disorders – Other rare conditions with some similar features to SPS include hyperekplexia, which presents in infancy, and psychogenic movement disorder [60], which can generally be distinguished from SPS by physical examination. (See "Nonepileptic paroxysmal disorders in infancy", section on 'Hyperekplexia' and "Hyperkinetic movement disorders in children", section on 'Hyperekplexia' and "Functional movement disorders".)

POST-DIAGNOSTIC EVALUATION — The association of stiff-person syndrome (SPS, formerly called stiff-man syndrome) with type 1 diabetes and/or malignancy requires that the clinician be vigilant for these diagnoses. All patients with SPS should undergo routine cancer screening tests appropriate for their age. No other specific testing for malignancy is usually required.

An associated autoimmune disorder may be present in most patients with SPS. As part of the evaluation of patients diagnosed with SPS, we obtain the following studies:

●Complete blood count and differential white blood cell count

●Routine blood chemistries, including testing of renal function, liver chemistries, creatine kinase (or aldolase)

●C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR)

●Thyroid-stimulating hormone, thyroid peroxidase antibodies, and hemoglobin A1C (glycated hemoglobin)

We do not perform specific autoantibody testing (eg, antinuclear antibody, rheumatoid factor, anti-cyclic citrullinated peptide) unless there are suspicious clinical features to warrant such testing.

TREATMENT — Treatment of stiff-person syndrome (SPS, formerly called stiff-man syndrome) is directed at the control of symptoms to improve mobility and function (algorithm 1).

Initial/symptomatic therapy — Benzodiazepines are generally considered the optimal initial therapy for patients with SPS. For initial symptomatic treatment in patients with SPS, we suggest a benzodiazepine (eg, diazepam 20 to 80 mg/day in three or four divided oral doses or clonazepam 1 to 3 mg orally two to four times daily). For example, diazepam can be started at an initial dose of 5 mg four times daily (taken orally) with a gradual upward titration of the drug over the course of several weeks to achieve a dose that is adequate to control the stiffness symptoms and allow for better mobility. This dose of diazepam can reach levels as high as 30 mg four times daily. By dividing the dose, patients may experience more sustained control of their symptoms. If diazepam is not well-tolerated, one can use clonazepam, beginning at a dose of 1 to 3 mg four times daily that can be titrated upwards to a maximum of 6 to 8 mg four times daily. Daily doses of diazepam needed to control the disease can be as high as 100 to 200 mg/day in some patients. We generally start with 5 mg two to three times daily and gradually increase the dose as tolerated and as required for therapeutic effect over several weeks. The dose response should be evident within a few weeks of starting the benzodiazepine and adjusted accordingly. If no improvement is noted with high doses such as 60 mg per day, the drug should be stopped and replaced with shorter-acting drugs such as oxazepam or lorazepam in equivalent doses. Major adverse effects that may limit further dose increases include central nervous system (CNS) symptoms, particularly somnolence and fatigue; mood changes and depression can also occur.

There are no randomized trials to guide the choice of initial drug, but the observation in case reports and case series that diazepam was effective for SPS provided a major breakthrough in the management of this previously untreatable disorder [7,35,61,62]. As an example, in one series sustained improvement was reported in most patients treated with a benzodiazepine, usually diazepam (median dose 40 mg daily in anti-glutamic acid decarboxylase [GAD] antibody seropositive classic SPS, and 17.5 mg daily in seropositive partial SPS) [35]. Benzodiazepines are thought to affect the condition by modulating the levels and activity of gamma amino butyric acid (GABA).

Resistant to initial symptomatic therapy — In patients with an unsatisfactory response or intolerance of escalating doses of benzodiazepines, we suggest baclofen (administered orally starting at 10 mg two to three times daily and gradually titrated upwards based upon tolerability and responsiveness as needed to a maximum of 80 mg daily in divided doses) rather than other muscle relaxants. We either switch to baclofen, a GABA-modulating drug, as monotherapy or add baclofen to the benzodiazepine, depending upon whether the patient has had some benefit from the dose of benzodiazepine that is tolerated. A response to baclofen is typically noted within a few weeks. Caution must be exercised when using these drugs in combination in order to avoid the commonest side effect which is sedation.

Baclofen has shown benefit in case reports and case series, and its use is supported by clinical experience, but it has not been tested in randomized trials [3,4,35]. Baclofen has occasionally also been given in the past by an intrathecal route, but we generally no longer use this route of administration with the availability of other safer alternatives, such as intravenous immune globulin (IVIG), with comparable efficacy (see 'IVIG' below). Complications of the procedure included spasm-induced rupture and dislocation of the intrathecal catheter [63,64].

Severe disease/resistance to symptomatic therapies — In patients with inadequate symptomatic relief from a positive allosteric modulator of the GABA receptor (eg, benzodiazepines), GABA receptor agonists (eg, baclofen), or other agents and in those with severe disease (symptoms that continue to significantly interfere with daily function and activities), we use immune modulating therapy. We generally initiate such treatment with IVIG therapy (see 'IVIG' below). Patients unresponsive to IVIG may respond to other therapies, such as plasma exchange or B-cell depletion with rituximab. (See 'Resistant to IVIG' below.)

IVIG — In patients not responsive to or intolerant of benzodiazepines as monotherapy or in combination with baclofen, if tolerated, we suggest IVIG rather than glucocorticoids or other immunosuppressive or immunomodulatory agents. IVIG is administered initially as a total dose of 2 g/kg, usually divided into two consecutive daily doses of 1 g/kg, or as 0.4 g/kg over five consecutive days [65-67]. The use and adverse effects of IVIG are described in detail separately. (See "Overview of intravenous immune globulin (IVIG) therapy" and "Intravenous immune globulin: Adverse effects".)

In patients who do not respond to the first round of IVIG with any significant change in function, an additional one or two rounds of IVIG can be tried, with dosing of 2 g/kg per infusion given every two weeks for up to six doses. For patients failing this course of treatment, rituximab or another intervention would be the next option. (See 'Resistant to IVIG' below.)

The frequency of follow-up treatment courses is highly variable and depends upon the duration of the response. Some patients may show sustained improvement for several months following their initial course of IVIG, whereas others require frequent infusions of 0.5 to 1.0 g/kg that are determined based upon their clinical response.

Evidence supporting the use of IVIG in patients with SPS includes a small randomized crossover trial and some case reports and case series; even severely affected patients may respond to this therapy [65-68]. The efficacy of IVIG was best shown in the randomized blinded crossover trial comparing IVIG with placebo [68]. Eleven of 16 patients who received immune globulin were able to walk more easily or without assistance; their frequency of falls decreased, and they were able to perform work-related or household tasks. The duration of benefit ranged from six weeks to one year. Anti-GAD65 antibody titers declined after therapy with IVIG but not after the administration of placebo. In another study, functional status improved markedly in three bedridden patients with previously refractory disease [65].

Resistant to IVIG — Several different options may be tried in patients resistant to treatment with IVIG, including B-cell depletion with anti-CD20 (rituximab), plasma exchange, and immunosuppressive agents, but studies are very limited, and responses have varied substantially between patients.

●B-cell depletion – B lymphocyte depletion and associated diminution of autoantibody levels may be beneficial in SPS. For patients who do not respond adequately to IVIG, rituximab, the chimeric B cell-depleting anti-CD20 monoclonal antibody, can be administered at a dose of 375 mg/m² given weekly for four consecutive doses. This regimen can be repeated every four to eight months, depending upon the patient's clinical course.

Benefit with rituximab has been described in some case reports [69,70] but not in others [71,72]. A placebo-controlled randomized trial of rituximab (two biweekly infusions of 1 gram each) studied 24 patients with SPS [73]. No differences were noted at six months in the stiffness index, the primary outcome, or in changes in heightened sensitivity scores at three or six months. Quality-of-life scores improved at three months in both groups but not at six months, suggesting an early placebo effect. Blinded self-assessment rating of the overall stiffness for individual patients revealed improvement in four patients in each group. At six months, improvement persisted in one patient in the placebo group compared with three out of four in the rituximab group, where these improvements were also captured by video recordings.

In another report, a patient with disease that was refractory to other treatments had dramatic clinical improvement following infusion of rituximab [69]. Some experts have suggested that patients with SPS and anti-GAD or anti-amphiphysin antibodies who are unresponsive to immunotherapy (including glucocorticoids and IVIG), as well as patients with progressive encephalomyelitis with rigidity and myoclonus (PERM), may benefit from a trial of rituximab [3]. Further study of rituximab is needed to better determine its effectiveness.

We prefer rituximab over plasma exchange because of the greater cost, the short-term nature of its benefit, and the need for ongoing intravenous access with the latter.

●Plasma exchange – Some patients who have been unresponsive to first-line therapies, including IVIG, as well as immunosuppressives, have been reported to respond to treatment with plasma exchange (plasmapheresis), which, for patients without a sufficient response to rituximab, may be used as the next treatment option. Roughly half of those patients with SPS treated with this modality who were described in case reports and case series experienced clinically significant improvement with this treatment [74]. Patients with SPS and life-threatening respiratory decline have also been treated with plasma exchange with variable results [75,76]. The short-term benefit, along with the high cost of the procedure, precludes its widespread use.

●Other therapies – A number of other potential therapies have been tried in individual cases and small numbers of patients, but the role of these agents and the expected degree of benefit relative to other therapies is unclear. These include the muscle relaxant methocarbamol given intravenously [77], botulinum toxin A (injected into the paraspinal muscles) [78], sodium valproate [79], vigabatrin [80], and propofol [81].

One report has also described responses to treatment with the calcineurin inhibitor, tacrolimus, in two patients who relapsed after initial benefit from IVIG or plasma exchange [82].

Glucocorticoids were used in the past, but we avoid their use in SPS because of the observations that long-term, high-dose therapy, with the resultant associated adverse effects of such therapy, is often required for patients to experience clinical improvement. In addition, glucocorticoid therapy should be avoided whenever possible in patients who also have diabetes because of the risk of adverse effects associated with their use in these patients. Other agents that are no longer used for SPS include azathioprine, methotrexate, and mycophenolate mofetil.

Patients are monitored for clinical changes or decline in function. Repeat GAD antibody testing is not considered useful since there is limited correlation between antibody titers and disease activity [52].

PROGNOSIS — The prognosis for patients with stiff-person syndrome (SPS, formerly called stiff-man syndrome) is guarded. Patients with more limited disease may continue to function normally for extended periods of time [1]. However, a functional decline is generally noted in most patients. Gait decline is inevitable for many patients. For those with partial SPS/stiff-limb syndrome, there may be significant loss of function that begins immediately. Thus, prompt initiation of treatment in an effort to maintain or improve functional status should be the goal for all patients.

A long-term follow-up of 57 patients with SPS demonstrated disease progression leading to physical disability; 46 patients (80 percent) lost the ability to walk independently despite treatment with symptomatic medications [83]. Over time, the number of stiff areas increased, consistent with worsening functional status and quality of life. High-titer anti-glutamic acid decarboxylase (GAD) antibodies were present in serum and cerebrospinal fluid (CSF) with elevated intrathecal GAD-specific IgG synthesis, but they did not correlate with clinical severity or progression.

With an effective combination of therapies, many patients can maintain reasonable levels of activity, though over time the majority of patients become disabled. When patients begin to lose function, a careful review of their medication regimen may identify therapeutic opportunities. For example, titration of benzodiazepines may require careful dosing adjustments; in our experience, most patients require taking their first dose upon arising, followed by three additional doses spaced at intervals throughout the rest of the day.

In patients who are left untreated, many will lose mobility due to the effects of persistent muscle spasm causing irreversible large joint contractures, especially at the hip, knee, ankle, and shoulder. Once these have developed, they are extremely difficult to overcome with drug therapy. In selected patients, surgical procedures, such as a Z-plasty procedure at the ankle joint, may partially correct a plantar flexion deformity at the ankle joint which, by itself, can severely limit ambulation.

Patients with paraneoplastic SPS may see marked improvement or complete remission of SPS following successful resection of their tumor [12].

SUMMARY AND RECOMMENDATIONS

●Stiff-person syndrome (SPS, formerly called stiff-man syndrome) is an uncommon disorder characterized by progressive muscle stiffness, rigidity, and spasm involving the axial muscles, resulting in severely impaired ambulation. SPS is often associated with type 1 diabetes mellitus (T1DM) and, less often, with thyroiditis, vitiligo, and pernicious anemia. Antibodies in the cerebrospinal fluid (CSF) of the majority of patients with SPS target GABAergic (gamma amino butyric acid) neurons and their nerve terminals. The dominant antigen recognized by these antibodies is glutamic acid decarboxylase (GAD). (See 'Introduction' above and 'Pathogenesis' above.)

●Patients with SPS have been subdivided into three categories, which include classic SPS in 70 to 80 percent of patients with SPS, with truncal stiffness, generalized rigidity, and frequent muscle spasms, resulting in an awkward, wide-based gait; partial SPS in 10 to 15 percent, most often presenting as a stiff-limb syndrome in which patients have marked difficulty with ambulation due to stiffness and lack of mobility affecting one limb, generally a leg, although other localized forms may occur; and other more uncommon forms, including a paraneoplastic SPS variant and progressive encephalomyelitis with rigidity and myoclonus (PERM). (See 'Clinical subtypes of SPS' above and 'Classic SPS' above and 'Partial SPS' above and 'Paraneoplastic SPS' above and 'Progressive encephalomyelitis with rigidity and myoclonus' above.)

●Most patients present between the ages of 20 and 50, and women are affected two to three times more often than men. Signs and symptoms which are characteristic include a prodrome of stiffness and rigidity in axial muscles, slow progression of stiffness resulting in impairment of ambulation, fixed deformity of the spine and pronounced lordosis, and the presence of superimposed episodic spasms that are precipitated by sudden movement, noise, or emotional upset. Patients exhibit normal findings on motor and sensory nerve examinations, other than the abnormal gait and hyperreflexia, and exhibit normal intellect. (See 'Epidemiology' above and 'Clinical subtypes of SPS' above and 'Classic SPS' above and 'Partial SPS' above.)

●Paroxysmal autonomic dysfunction has been described and may result in sudden death, possibly related to sudden apnea and respiratory arrest from diaphragmatic spasm, impaired respiratory function, and severe respiratory muscle rigidity. Esophageal dysmotility and swallowing difficulties may lead to aspiration. Specific phobias may occur, and anxiety may further exacerbate muscle stiffness or may trigger spasms. Anxiety and avoidance behavior may lead to a misdiagnosis of a phobic disorder. (See 'Classic SPS' above.)

●Electromyographic (EMG) studies reveal continuous motor-unit activity which is typically decreased or abolished by intravenous diazepam, sleep, and local or general anesthesia. There are no distinguishing electrophysiologic features between GAD antibody-positive and -negative patients. No characteristic neuropathologic changes have been consistently identified. (See 'Electromyographic studies' above and 'Neuropathology' above.)

●The diagnosis is based upon the presence of characteristic clinical features. Muscle hardening to a board-like sensation is probably the most specific clinical observation. The presence of the following features are generally considered necessary for making the diagnosis:

•Stiffness in the axial and limb muscles resulting in impairment of ambulation

•Presence of superimposed episodic spasms that are precipitated by sudden movement, noise, or emotional upset

•A positive therapeutic response to oral diazepam or findings of continuous motor-unit activity on EMG that are abolished by intravenous diazepam

•Absence of other neurologic disorders that may explain the clinical features

When present, anti-GAD antibodies support the diagnosis; however, these antibodies may be absent in up to one-third of patients. In addition, in patients with characteristic clinical findings, anti-GAD antibodies, and a positive response to diazepam, the diagnosis can be made without an EMG, although such electrodiagnostic studies and a clinical response to diazepam can provide additional support for the diagnosis. (See 'Diagnosis' above.)

●The pain and stiffness described in SPS may mimic other conditions, including musculoskeletal and neurologic disorders. These include axial spondyloarthritis (AS), Parkinson disease (PD), tetanus, axial dystonia, psychogenic movement disorder, hyperekplexia, and focal limb or joint disorders. (See 'Differential diagnosis' above.)

●Post-diagnostic testing includes routine cancer screening tests appropriate for age; a complete blood count and differential white blood cell count; routine blood chemistries, including testing of renal function, liver chemistries, creatine kinase (or aldolase); C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR); and thyroid-stimulating hormone, thyroid peroxidase antibodies, and hemoglobin A1C (glycated hemoglobin). There should be a high level of suspicion for type 1 diabetes or other autoimmune disease. (See 'Post-diagnostic evaluation' above.)

●Treatment of SPS is directed at the control of symptoms to improve mobility and function (algorithm 1). For initial symptomatic treatment in patients with SPS we recommend a benzodiazepine (eg, diazepam 20 to 80 mg/day in three or four divided oral doses or clonazepam 1 to 3 mg orally two to four times daily) rather than baclofen or an immunosuppressive agent (Grade 2C). Daily doses of diazepam needed to control the disease can be as high as 100 to 200 mg/day in some patients. We generally start with 5 mg two to three times daily and gradually increase the dose as tolerated and as required for therapeutic effect over several weeks. (See 'Initial/symptomatic therapy' above.)

●In patients with an unsatisfactory response or intolerance of escalating doses of benzodiazepines, we suggest baclofen rather than other muscle relaxants or immunosuppressives (Grade 2C). Baclofen is usually started at a dose of 10 mg two to three times daily and gradually titrated upwards as tolerated and needed to a maximum of 80 mg daily in divided doses. We either switch to baclofen, a GABA-modulating drug, as monotherapy or add baclofen to the benzodiazepine, depending upon whether the patient has had some benefit from the dose of benzodiazepine that is tolerated. (See 'Resistant to initial symptomatic therapy' above.)

●In patients not responsive to or intolerant of benzodiazepines as monotherapy or in combination with baclofen, if tolerated, we suggest intravenous immune globulin (IVIG) rather than glucocorticoids or other immunosuppressive or immunomodulatory agents (Grade 2C). IVIG is administered initially as a total dose of 2 g/kg, usually divided into two consecutive daily doses of 1g/kg, or as 0.4 g/kg over five consecutive days. (See 'Severe disease/resistance to symptomatic therapies' above and 'IVIG' above.)

●Several different options may be tried in patients resistant to treatment with IVIG, including B cell depletion with anti-CD20 (rituximab), plasma exchange, and immunosuppressive agents, but studies are very limited, and responses have varied substantially between patients. (See 'Resistant to IVIG' above.)

●The prognosis for patients with SPS is variable. Patients with more limited disease may continue to function normally for extended periods, but gradual functional decline is generally noted. (See 'Prognosis' above.)