Clinical Spectrum of Stiff Person Syndrome: A Review of Recent Reports

Abstract

Background

“Classic” stiff person syndrome (SPS) features stiffness, anti-glutamic acid decarboxylase (anti-GAD) antibodies, and other findings. Anti-GAD antibodies are also detected in some neurological syndromes (such as ataxia) in which stiffness is inconsistently present. Patients with otherwise “classic” SPS may either lack anti-GAD antibodies or be seropositive for others. Hence, SPS cases appear to fall within a clinical spectrum that includes conditions such as progressive encephalomyelitis with rigidity and myoclonus (PERM), which exhibits brainstem and autonomic features. We have compiled herein SPS-spectrum cases reported since 2010, and have segregated them on the basis of likely disease mechanism (autoimmune, paraneoplastic, or cryptogenic) for analysis.

Methods

The phrases “stiff person syndrome”, “PERM”, “anti-GAD antibody syndrome”, and “glycine receptor antibody neurological disorders” were searched for in PubMed in January 2015. The results were narrowed to 72 citations after excluding non-English and duplicate reports. Clinical descriptions, laboratory data, management, and outcomes were categorized, tabulated, and analyzed.

Results

Sixty-nine autoimmune, 19 paraneoplastic, and 13 cryptogenic SPS-spectrum cases were identified. SPS was the predominant diagnosis among the groups. Roughly two-thirds of autoimmune and paraneoplastic cases were female. Anti-GAD antibodies were most frequently identified, followed by anti-amphiphysin among paraneoplastic cases and by anti-glycine receptor antibodies among autoimmune cases. Benzodiazepines were the most commonly used medications. Prognosis seemed best for cryptogenic cases; malignancy worsened that of paraneoplastic cases.

Discussion

Grouping SPS-spectrum cases by pathophysiology provided insights into work-up, treatment, and prognosis. Ample phenotypic and serologic variations are present within the categories. Ruling out malignancy and autoimmunity is appropriate for suspected SPS-spectrum cases.

Introduction

“Stiff man” syndrome was first described in 1956 by Moersch and Woltman.^1,2 Along with observations from 13 other cases, they described a 49-year-old man with progressive stiffness in his neck, shoulders, and upper back, episodic painful muscle spasms, and difficulty walking. Multiple similar case descriptions have since followed. The term “stiff man” was recently replaced by the gender-neutral “stiff person syndrome” (SPS), which gained significant traction after Blum and Jankovic³ reported that approximately 20 of the 84 reported cases between 1967 and 1991 were female. It was Asher,⁴ however, in 1958, who first proposed this terminology.

The suspicion for an immunologic cause was raised by the observations of frequent comorbid diabetes (up to 35% in some series⁵) and other concomitant autoimmune diseases (vitiligo, celiac sprue, rheumatologic diseases, and thyrogastric disorders)^2,5,6 in patients with SPS. Glutamic acid decarboxylase (GAD) antibodies (which in this manuscript will be referred to as anti-GAD antibodies, a non-specific term that includes both anti-GAD antibody isoforms, as described below) were first documented in association with SPS in 1988.^7,8 Anti-GAD antibodies inhibited GAD activity and the synthesis of gamma-aminobutyric acid (GABA) in vitro.²

GAD is a pyridoxal 5′-phosphate-dependent enzyme and the rate-limiting step in the synthesis of GABA. GAD is not only found in the brain and pancreatic B-cells, but also in lower amounts in the liver, kidneys, adrenal glands, ovaries, and testes.⁹ There are two GAD isoforms, 65 and 67, which differ in their molecular weight, location, and enzyme activity. Within the central nervous system, GAD65 localizes to the synaptic vesicles and its activity increases in response to surging demands for GABA.² GAD67 localizes to the cytoplasm and generates a steady basal GABA level.² Anti-GAD antibodies are specific for either isoform: antibodies against GAD65 were reported in about 80% of SPS cases (at times, the terms anti-GAD and anti-GAD65 antibodies are used interchangeably in the literature), whereas anti-GAD67 antibodies were reported in about 60%, with co-existence presumed likely.^2,6,10

An immune pathogenesis is accepted as the cause of SPS, but it remains unclear whether anti-GAD antibodies are directly pathogenic in vivo, unlike in diabetes.² There are different possible explanations for this. Whereas serum anti-GAD antibody titers in SPS are high enough to produce endocrine damage, diabetics have lower serum titers that are likely insufficient to cross the blood–brain barrier and lead to central nervous system (CNS) damage.² Extremely high titers of anti-GAD antibodies in SPS can trigger multi-antigen autoimmunity and the development of other concomitant autoimmune diseases, such as thyroid disease, but not necessarily vice versa.⁹ Besides a difference in titers, in SPS both the linear and the conformational epitopes are recognized by anti-GAD65 antibodies, while in diabetes only the conformational epitopes are recognized, triggering a specific pathogenic mechanism.² Other arguments against a pathogenic role for these antibodies include the lack of correlation between disease severity and intrathecal synthesis and lack of evidence of T-cell-mediated damage of central GABAergic neurons.^5,10 In addition, antibody titers in SPS do not appear to vary with clinical response to treatment. Monitoring antibody titers during the course of treatment is, hence, unnecessary.^5,11

Identification of the anti-GAD antibodies allowed the revision of the diagnostic criteria for SPS, which were first proposed in 1967 by Gordon et al.¹² In 2009 Dalakas⁵ proposed two fundamental clinical symptoms, truncal and proximal limb stiffness, stemming from co-contraction of agonist and antagonist muscles and leading to hyperlordosis, and superimposed episodic spasms. According to Dalakas, the presence of all of the following are required for a diagnosis of SPS: 1) stiffness of the axial muscles, particularly in the abdomen and thoracolumbar paraspinals, leading to hyperlordosis; 2) superimposed painful spasms triggered by tactile or auditory stimuli; 3) electromyographic evidence of continuous motor unit activity in agonist and antagonist muscles; 4) absence of other neurological findings that may suggest an alternative diagnosis; and 5) positive serology confirmed by immunocytochemistry, Western blot, or radioimmunoassay.

Although these criteria best define the “classic” SPS phenotype, it is now clear that some patients have positive anti-GAD antibodies and stiffness that is confined to a limb (typically a leg, in what has been called stiff-limb syndrome, SLS).¹³ Eventual spread of stiffness to the trunk was described.¹³ SLS is likely due to local interneuronitis, in which there is selective destruction of spinal interneurons in the gray matter.¹³ Interestingly, long-tract damage was not noted in these cases. The associated reflexive spasms likely result from an excessive response to descending reticulospinal activity at the segmental level, as shown electrophysiologically by hypersynchronous segmented discharges.¹³ The exact cause of the interneuronitis is unknown but may be due to similar mechanisms that occur in the presence of intrinsic spinal cord tumors or vascular insufficiency.¹³ These patients may develop urinary and transient brainstem symptoms (about 50%), have seropositivity to anti-GAD antibodies in about 15% of cases, and have a limited response to GABAergic treatments.^13,14 In fact, patients with stiff limb syndrome were reported likely to be wheelchair bound after a mean of 3.5 years.¹³ Some authors also advocate for the existence of a distinct stiff-trunk syndrome,¹⁴ and a jerking-man syndrome with generalized myoclonus.^13,15 The development of highly sensitive antibody detection assays also allowed the identification of anti-GAD antibodies in neurologic conditions where rigidity may be absent. Phenotypes include cerebellar ataxia, epilepsy, and cognitive impairment.² All of the above suggest the existence of a spectrum of SPS expanding well beyond the “classic” phenotype.

Patients within this SPS spectrum have antibodies against other proteins of the GABAergic synapse, including amphiphysin and gephyrin, which may be identified in isolation.^10,16 Amphiphysin is a cytosolic pre-synaptic vesicle protein. Anti-amphiphysin antibodies are often associated with malignancy, and phenotypically these cases may differ from anti-GAD65 positive SPS with more prominent neck and arm stiffness.¹⁷ In addition, these antibodies were also detected in neurologic conditions such as encephalopathy, myelopathy, and neuronopathy.¹⁷ Gephyrin interacts with the GABA receptor-associated protein (GABARAP) in the assembly of the GABA-A receptor.^10,16 Interestingly, about 70% of patients with anti-GAD65 seropositivity may also have antibodies against GABARAP.^10,16

Antiglycine receptors (anti-GlyR antibodies) are present in some SPS-spectrum cases, and are the hallmark of progressive encephalomyelitis with rigidity and myoclonus (PERM). In this condition there is prominent brainstem, autonomic, and spinal cord involvement. Concurrent anti-GAD antibodies can also be found.¹⁸ In a study of 45 prospective and 56 retrospective cases of PERM, approximately 33 of the former and 10 of the latter were positive for anti-GAD antibodies.¹⁸ Histology from available specimens demonstrated inflammatory and microglial changes and cell loss in the pons, medulla, cerebellum, spinal cord, and autonomic ganglia.¹³ Patients with PERM had increased T2 fluid-attenuated inversion recovery signal of spinal cord and brainstem on magnetic resonance imaging (MRI).¹⁸ Despite its severity, the response to immunomodulatory therapies, including methylprednisolone, can be robust.^13,18

Glycine receptors belong to a family of ligand-gated ion channels composed of two alpha- and three beta-subunits. Activation of these receptors leads to chloride influx, membrane hyperpolarization, and reduction in neuronal excitation. There are three types of glycine receptors: alpha 1, 2, and 3. The first type localizes to the brainstem, thalamus, hypothalamus, superior colliculus, and spinal cord. The distribution of the others remains incompletely defined.¹⁸

Other antibodies were found in patients within the spectrum of SPS. In a study examining the antibody profile of 13 patients with SPS, eight were found to have non-organ-specific antibodies, including antibodies to nuclear, smooth muscle, and mitochondrial antigens, six had thyrogastric antibodies, five had islet-cell antibodies, and four had anti-GAD antibodies.⁹

Given the considerable phenotypical variability found in patients with anti-GAD antibodies, the presence of different auto-antibodies associated with similar phenotypes, and seronegative patients fulfilling almost all of Dalakas’ criteria for “classic” SPS, classifying patients within the SPS spectrum-based on their phenomenology can be, in our opinion, impractical and ambiguous. Instead, we argue that a classification based on likely etiology offers the most useful guidance in terms of prognosis and treatment response. Under this framework, cases within the SPS spectrum can be segregated into one of three mutually exclusive groups:^9,19,20 1) autoimmune cases, defined by autoantibody positivity (in serum and/or cerebrospinal fluid [CSF]) in the absence of an underlying malignancy; 2) paraneoplastic cases, encompassing all cases emerging in the context of cancer; and 3) cryptogenic cases, that is, all seronegative cases in which an immunologic cause cannot be identified. All cases with malignancy were labeled as paraneoplastic, as the original reports described a mostly positive clinical response to anti-cancer treatments. If the cancer therapy did not produce a response in the SPS-spectrum disorders, the malignancy would more likely be a comorbidity rather than a likely cause. This breakdown is purely an operational classification based on the published papers’ reported diagnosis. It was difficult for us to determine if each case met Dalakas’ criteria in their entirety, as phenotypic descriptions were incomplete in a number of the reports, despite the reported diagnoses. To better understand the characteristics, disease behavior, and treatment response of SPS-spectrum cases within each of these groups, we here review cases and case series of SPS and related disorders published between 2010 and early 2015.

Methods

In January 2015, we searched PubMed for the phrases “stiff person syndrome”, “PERM”, “anti-GAD antibody syndrome” and “glycine receptor antibody neurological disorders”. Of note, searching for “stiff man syndrome” yielded the same results as searching for “stiff person syndrome.” The resulting initial 706 citations were narrowed to 72 after excluding non-English papers, duplicate reports, and papers published before 2010. Papers with no available clinical information and reviews without individual clinical descriptions were also excluded. January 2010 was selected as a cut-off for inclusion in an attempt to guarantee that most of the published cases were using similar criteria to diagnose SPS-spectrum cases, but also to adhere to the journal’s manuscript guidelines. Clinical descriptions, laboratory data, treatment strategies, and outcomes, when available, were extracted and tabulated into a Microsoft® Excel 2010 spreadsheet. Three separate spreadsheets were created, one for each subgroup (autoimmune, paraneoplastic, or cyptogenic), and cases were segregated into one of these categories according to the definitions mentioned above. Case series were treated as groups of case reports, and each of the cases within the series was classified independently into one of the tables. One case series²¹ presented consolidated data for all of its cases and is presented here in a separate table (see Table 4) that includes autoimmune, paraneoplastic, and cryptogenic subtypes. Attempting to subdivide these cases into the prior three subgroups was thus not readily possible due to lack of individual data.

Table 4. Reported Cases of Cryptogenic Stiff Person Syndrome and Its Variants.

				Symptoms									History or Examination Findings											Testing								Treatment						Outcome
Reference	Reported Diagnosis	Age And Sex at Presentation	Number Of Symptomatic Months Before Presentation	Bulbar Dysfunction	Genitourinary Dysfunction	Difficulty Moving Limb	Falls	Gait Difficulties (Including Ataxia)	Pain	Axial Stiffness Or Hyperlordosis	Autonomic Instability (Besides Genitourinary Dysfunction)	Concomitant Autoimmunity	Cramps/Spasms	Encephalopathy Or Psychiatric/Behavioral Changes	Hyperekplexia	Hyperreflexia	Limb Posturing/Dystonia	Limb Stiffness/Rigidity	Myoclonus	Seizures	Weakness	Documentation Of Autoantibody Testing	Serum Anti–Gad Or Anti–Glyr Antibodies	CSF Oligoclonal Bands	Electroencephalographic Abnormalities	Electromyography Findings Consistent With SP	MRI Brain Abnormalities (Incidental Findings Excluded)	Antiepileptics	Baclofen	Benzodiazepines	Botulinum Toxin Injections	Intravenous Immune Globulin	Other agents or interventions	Plasmapheresis	Steroids	Steroid–sparing immunosuppression	Surgical intervention	Improvement or resolution	1 or more relapses	Stabilization without improvement
Ughratdar et al.32	SL, SP	44M	168						+				+		+		+					+	–			+			+	+			+				+	+
Hegyi.40	SP	24F	9			+		+	+	–			+	+			+	+			+	–								+			+					+
Iwata et al.34	SL	29M	24					+	+				+		+	+	+	+				+	–	+	–	+	–			+			+					+
Prasad89	SP	51M	6				+		+		–		+	+	+					+	+	–			–		–		+	+			+							+
Sanefuji et al.90	SP	7F	<1					+	+				+	+	+	–			–			+	–		–	+	+		+	+		+	+		+			+
Schreiber et al.41	SP	75F	36				+	+	+	+			+	–	+	+		+				+	–			–	–		+	+			+					+
Lorenzoni et al.33	SP	43M						+	+	+		T, V	+		+	+		+				+	–			+			+	+								+
Clardy et al.14	SP	13M	2							+		–	+			–						+	–										+	+	+			+
Newton et al.42	SP	48M		+				+	+	+			+			–		+				+	–						+	+	+							+
Vicente–Valor et al.43	SP	40M	72			+		+	+	+	+		+		+	+		+				+	–			+		+	+	+		+	+		+	+		+
Sharma et al. 91	SP	65M	6					+	+	+			+	+		–		+				+	–			+			+	+					+			+
Pakeerappa et al.92	SP	48M	17							+			+					+				+	–						+	+	+		+					+
Ueno et al.35	PERM	48F	<1		+		+	+	+	+	+		+	+		+		+	+			+	–	–		+	+	+		+		+	+		+	+		+	+

Pertinent positives and pertinent negatives for symptoms, examination findings, laboratory data, treatment strategies, and outcomes were coded in the tables as present (+) or absent (–), respectively. Fields were left blank when the paper did not comment on a specific datum. Once all data were entered, findings were summated within each category and percentages of prevalence were calculated. Either medians or means were calculated for all numeric data (i.e. age), depending on the sample size. The prevalence numbers for the subdivisions are based on the published diagnoses of SPS and its variants in the original reports.

Results

Data extracted from each of the SPS-spectrum subgroups are presented in the following subsections.

Autoimmune SPS spectrum cases (Tables 1 and 2)

Table 1. Subjective and Objective Findings in Reported Cases of Autoimmune Stiff Person Syndrome and Its Variants.

				Symptoms									History or Examination Findings																Antibody Testing							Other Testing
Reference	Reported Diagnoses at Presentation	Age and Sex at Presentation	Number Of Symptomatic Months Before Presentation	Bulbar (Excluding Taste Changes)	Taste Changes	Genitourinary Dysfunction	Difficulty Changing Position	Difficulty Moving Limb	Falls	Gait Difficulties (Including Ataxia)	Pain	Pruritus	Axial Stiffness or Hyperlordosis	Autonomic Instability (Besides Genitourinary Dysfunction)	Concomitant Autoimmunity (Other Than Diabetes)	Cramps/Spasms	Diabetes	Extraocular Movement Abnormalities	Hepatitis C	Encephalopathy or Psychiatric/Behavioral Changes	Hyperekplexia	Hyperreflexia	Limb Posturing/Dystonia	Limb Stiffness/Rigidity	Malignancy	Myoclonus	Seizures	Weakness	Serum Anti-GAD Antibodies	Serum Anti-Glyr Antibodies	Other Serum Antibodies Detected	CSF Anti-GAD Antibodies	CSF Anti-Glyr Antibodies	Other CSF Antibodies Detected	CSF Oligoclonal Bands	EEG (Findings Consistent With Seizures)	Electromyography Findings Consistent With SP	MRI Brain Abnormalities	MRI Spine Abnormalities To Explain The Patient’s Symptoms
Lobo et al.44	SP	41F	84				+				+		+			+	–							+	–				+								+
Scavone et al.24	SP	66M								+			+			+	+		+	–	+		+	+				–	+		AT, AG
Awad et al.23	A	48F		+						+					P		–	+	+										+		AI, ANA, AP, SSA	+			+			–	–
Castelnovo et al.45	SL	63F	36							+	+		–		P	+	+				+		+	+					+						–		–	–	–
Cuturic et al.46	SP, ED	35F	24										+			+	+			+				+		–			+							–	–	–
Ehler et al.6	SP	61M	<1	+		+		+					+	+		+	+							+					+		K			K		–	+	–	–
Gnanapavan et al.47	SP	45M	60						+	+			+			+	+					+							+								+
Goldkamp et al.48	SP	27F														+													+
Mas et al.37	E, PERM	60M	<1	+							+			+		+		+		+	+	+		+		+	+	+	–	+	–		+		–		+	–
	PERM	48M	2		+							+	+	+		+				+	+	+		+					–	+	–		+		–		–	–	–
	SL, SP	33F	3	+		+				+	+				T	+		+					+	+		+			–	+	–				–		+	–	–
Piotrowicz et al.49	PERM	58M	>1						+	+			+					+			+	+		+	–	+		+	–	+	–		+		+		+	–
Turner et al.27	PERM	28M	1	+		+								+				+		+	+	+		+		+	+		–	+	AN					+		–	–
Witherick et al.28	SP	69M									+				MG, P	+					+			+					+								+
Anagnostou et al.50	SL	40F	108							+	+					+	+				+		+	+					+								+	–	–
Amyradakis et al.51	SP	F	<1	+							+					+	+	+						+					+
Fekete and Jankovic52	SP	12M	84							+	+		+	+		+					+	+		+		+	+		+		–					–		–
Fernandes et al.53	SP	50F	48						+	+	+		+			+									–				+								+	–	–
	A, E	52F	<1	+						+								+									+		+		–					–		–
Iizuka et al.25	PERM	61F	1.5	+	+					+						+		+				+		+		+			+	+	AG, AX	+	+		+		+	–
Lorenzoni et al.33	SL	10F								+	+			+		+						+		+					+								+
	SP	40M								+	+					+	+							+					+								+
	SP	42M								+	+		+			+	+							+					+								+
Najjar, at al.22	S	19F															–			+									+		–					–		–
Peeters et al.36	PERM	37F	1	+		+			+		+	+	+	+		+		+		–	+	+		+	–	+			–	+	–		+		+		+	–	–
Qureshi et al.54	SP	56M	72	+							+		+			+	+	–		+		–		+		+		+	+		–				+	–		–	–
Tsai et al.55	SP	66M	4	+					+							+	+				+	+	+	+	–a	+		+	+		–				–			–
Baroncini et al.26	A, LE	44F	36	+		+				+			+		T	+		+		+			+	+	–		+		+		AP, AX	+			+	–	+	+
Clardy et al.14	SP	8F	<12												V		+					–		+					+	+							–
	SP	26M	168	+									+		V							+		+					+	–
	SP	51F	552										+		–							–		+					+	–			+				–
	SP	49M	528	+											G		+					–		+					+	–							–
	SL	14F	156												T		+					+		+					+	–			–				–
	ST	17M	36										+		–							–							+	–			+				+
	PERM	13F	<12										+			+	+				+	+		+					+	–							–
De la Casa-Fages et al.56	SP	59F	10	+						+			+		P, V	+								+					+		–						+
	SP	48M	240						+	+			+		V	+	+				+			+					+		–						+
Damasio et al.57	PERM	1F	<1	+		+				+			+	+		+		+		+	+	+	+	+		+				+	–		+			–		–
Marinovic et al.58	SP	51F								+	+		+			+													+		–						+		–
Nakane et al.59	SP																												+
	SP																												+
O’Toole et al.60	SP	72F	12							+			+		T	+	+				+			+	–				+								+
Sidransky et al.61	SP	34	6	–							+		+								+								+
Sengupta et al.62	A, SP	F																											+
Vetrugno et al. 201363	OT, SP	77F	1						+	+	+		+			+					+			+	–				+		–							–	–
	OT, SP	55F	12						+	+	+		+		T, V	+					+			+	–				+		–
Bordelon et al.64	SP	60F	108						+				+	+		+				+		+		+					+							–	+	–
Enuh et al.65	SP	20F	12					+		+	+			+	T	+	+			+			+			+	–	+				+					+	–	–
Fourlanos et al.66	SP	78F	300										+			+	+			+	+	+		+					+			+						–	–
	SP	72F	6								+					+	+				+		+	+					+						–			–	–
Georgieva et al.67	A, E	45M		+						+							+	+		+		+					+		+			+			+			–
Ho et al.68	SP	43F		+				+	+	+			+			+		+		+	+	–						–	+		–					–	+	–	–
Jung et al.69	SP	55F	12						+	+	+					+		–			+	+		+				–	+								+
	SP	58F	15								+		+		P	+	+				+			+					+		–						–		–
	E, SP	49F	10										+									+		+			+	–	+							+	+	–
Pagano et al.30	SP	45M															+								–				+
	A, SP	65F															+								–				+
	SP	48F																	+						–				+
	SP	61F																							–				+		AA
	SP	34F															+								–				+
	A, SP	50F																							–				+
Rana et al.70	SP	50F	30	+							+		+		V	+													+
Sanders et al. 71	SP	48F							+	+										+	+	+		+				–	+								–	–	–
	SP	30F	24	+																	+			+					+								+
Stern et al.72	PERM	40M	<1	+										+		+		+		+	+			+	–	+			–	+	–		+		–			–
Wuerfel et al.73	B, E	2M	12	+										+						+	–		+	–	–	–	+		–	+	–		–		–	+		–
Bowen et al.29	C, SP	49F	24					+	+	+			+			+		–			+		–	+		–		–	+			+			+			–	–
	C, SP	68F	84					+	+	+			+				+	–						+					+						+			+	–
Farooqi et al.74	E	23F	48							–								–		–				–			+		+							+		+

Abbreviations: A, Ataxia; AA, Anti-amphiphysin Antibodies; AG, Anti-thyroglobulin; AI, Anti-intrinsic Factor; AN, Anti-NMDA Receptor Antibodies; ANA, Anti-nuclear Antibody; AP, Anti-parietal Cell Antibody; AT, Anti-thyroid Microsomal; AX, Anti-thyroid Peroxidase Antibody; B, Behavioral Changes; C, Corticobasal Syndrome; CSF, Cerebrospinal Fluid; E, epilepsy; ED, Eating Disorder; F, Female; G, Graves’ Disease; K, Antibodies against Tick-borne Meningoencephalitis; LE, Limbic Encephalitis; M, Male; MG, Myasthenia Gravis; MRI, Magnetic Resonance Imaging; OT, Orthostatic Tremor; P, Pernicious Anemia; PERM, Progressive Encephalomyelitis with Rigidity and Myoclonus; T, Autoimmune Thyroid Disease; S, Schizophrenia; SL, Stiff Limb Syndrome; SP, Stiff Person Syndrome; SSA, SS-A; ST, Stiff Trunk Syndrome; V, Vitiligo.

^aThis patient had no evidence of active malignancy at the time of symptom presentation; but had a remote history of treated lymphoma and thymoma.

Table 2. Treatments and Outcomes in Reported Cases of Autoimmune Stiff Person Syndrome and Its Variants.

				Treament													Outcome
Reference	Reported Diagnoses at Presentation	Age and Sex At Presentation	Number of Symptomatic Months Before Presentation	Antibiotics	Antiepileptics	Antipsychotics	Baclofen	Benzodiazepines	Botulinum Toxin Injections	Intravenous Immune Globulin	Other Agents Or Interventions	Plasmapheresis	Rituximab	Steroids	Steroid–Sparing Immunosuppression	Surgery: Spinal Cord Stimulator	Improvement or Resolution	Stabilization Without Improvement	One or More Relapses Mentioned	Worsening and/or Death
Lobo et al.44	SP	41F	84					+	+				+				+
Scavone et al.24	SP	66M								+							+
Awad et al.23	A	48F								+			+	+				+
Castelnovo et al.45	SL	63F	36					+		+							+
Cuturic et al.46	SP, ED	35F	24				+	+		+	+						+
Ehler et al. 6	SP	61M	<1				+	+				+		+	+		+		+
Gnanapavan et al.47	SP	45M	60				+	+		+	+		+		+		+		+
Goldkamp et al.48	SP	27F					+	+		+	+			+			+
Mas et al.37	E, PERM	60M	<1															+
	PERM	48M	2					+		+	+			+			+
	SL, SP	33F	3				+	+		+				+			+		+
Piotrowicz et al.49	PERM	58M	>1					+				+		+	+		+		+
Turner et al.27	PERM	28M	1		+															+
Witherick et al.28	SP	69M								+	+				+		+		+	+
Anagnostou et al.50	SL	40F	108					+	+								+
Amyradakis et al.51	SP	F	<1					+			+						+		+
Fekete and Jankovic52	SP	12M	84		+		+	+		+	+		+				+
Fernandes et al.53	SP	50F	48				+	+		+							+
	A, E	52F	<1		+			+		+				+	+
Iizuka et al.25	PERM	61F	1.5					+		+				+	+		+		+
Lorenzoni et al.33	SL	10F						+		+							+
	SP	40M					+	+		+							+
	SP	42M					+	+		+							+
Najjar, at al.22	S	19F				+				+		+		+			+		+
Peeters et al.36	PERM	37F	1				+	+				+		+			+		+
Qureshi et al.54	SP	56M	72		+			+		+	+	+	+	+			+
Tsai et al.55	66M	SP	4				+	+		+				+	+		+		+
Baroncini et al.26	A, LE	44F	36		+					+		+	+	+			+
Clardy et al.14	SP	8F	<12				+	+		+		+			+		+
	SP	26M	168				+	+						+				+
	SP	51F	552				+	+		+					+			+
	SP	49M	528															+
	SL	14F	156					+		+							+
	ST	17M	36				+	+	+		+	+					+
	PERM	13F	<12					+		+		+	+					+
De la Casa-Fages et al.56	SP	59F	10		+		+	+	+	+		+		+	+		+
	SP	48M	240					+		+		+		+	+		+
Damasio et al.57	PERM	1F	<1	+	+			+		+				+			+		+
Marinovic et al.58	SP	51F						+			+							+
Nakane et al.59	SP							+		+					+		+		+
	SP							+				+		+	+		+		+
O’Toole et al.60	SP	72F	12
Sidransky et al.61	SP	34	6				+	+		+	+						+
Sengupta et al.62	A, SP	F											+		+		+		+
Vetrugno et al.63	OT, SP	77F	1					+		+							+		+
	OT, SP	55F	12					+		+							+
Bordelon et al.64	SP	60F	108				+	+		+	+			+			+
Enuh et al.65	SP	20F	12				+	+		+	+						+
Fourlanos et al.66	SP	78F	300							+		+					+
	SP	72F	6				+			+						+	+
Georgieva et al.67	A, E	45M								+		+			+		+		+
Ho et al.68	SP	43F					+	+						+			+
Jung et al.69	SP	55F	12					+									+
	SP	58F	15				+	+									+
	E, SP	49F	10				+	+						+			+
Pagano et al.30	SP	45M										+					+
	A, SP	65F								+		+						+
	SP	48F								+		+					+
	SP	61F								+		+					+
	SP	34F								+		+					+
	A, SP	50F										+					+
Rana et al.70	SP	50F	30
Sanders et al. 71	SP	48F			+		+	+		+	+	+			+		+
	SP	30F	24				+	+		+	+	+					+
Stern et al.72	PERM	40M	<1				+	+		+		+		+			+		+
Wuerfel et al.73	B, E	2M	12		+									+			+
Bowen et al.29	C, SP	49F	24
	C, SP	68F	84							+										+
Farooqi et al.74	E	23F	48		+					+		+		+	+			+	+

Abbreviations: A, Ataxia; B, Behavioral Changes; C, Corticobasal Syndrome; E, Epilepsy; ED, Eating Disorder; F, Female; LE, Limbic Encephalitis; M, Male; OT, Orthostatic Tremor; PERM, Progressive Encephalomyelitis with Rigidity and Myoclonus; S, Schizophrenia; SL, Stiff Limb Syndrome; SP, Stiff Person Syndrome; ST, Stiff Trunk Syndrome.

Sixty-nine cases of autoimmune SPS-spectrum disorders were identified (Table 1), 46 (66.7%) of whom were female. Sex and age at presentation were available in all but four cases: two were female, but their age was not reported. Neither sex nor age was reported for the other two. Ages at presentation ranged from 1 to 78 years, the mean being 44 years. The median age at presentation in males was 45 years, whereas the mean age was 44.5 years in females.

In terms of their phenotypes, 48 cases (69.6%) received a diagnosis of SPS. Of these, three had concomitant ataxia, two had orthostatic tremor, two had corticobasal syndrome, one initially had stiff limb syndrome, one had epilepsy, and one had anorexia. Nine (13%) had findings consistent with PERM. Four (5.8%) received a diagnosis of SLS, and another four had ataxia alone. Two (2.9%) had only epilepsy, one of whom with behavioral changes. Finally, only one (1.4%) of the autoimmune cases had clinical signs of schizophrenia,²² while another had isolated stiff trunk syndrome.¹⁴

Data on the time patients had symptoms before presentation were available for 48. It ranged from less than 1 month to about 46 years, with an average of 63.8 months (over 5 years). Forty-one (59.4%) described cramps or spasms, 45 (65.2%) had limb stiffness, 29 (42%) had difficulty walking, and 23 (33.3%) had pain. Sixteen (23.2%) had documented concomitant autoimmunity besides diabetes (only two had documented absence of concomitant autoimmunity; data were lacking for the rest), and 26 (37.7%) had documented diabetes. In descending order of frequency, other associated autoimmune conditions included five with isolated thyroid disease; four with isolated vitiligo; three with isolated pernicious anemia; one with coexisting thyroid disease and vitiligo; one with coexisting pernicious anemia and vitiligo; one with coexisting myasthenia gravis and pernicious anemia; and one with Graves’ disease. Eight (11.6% of all autoimmune cases) had coexisting diabetes along with another autoimmune disease.

Serum anti-GAD antibody testing was documented for 66/69 (95.6%), 58/66 (87.9%) being seropositive. Twenty-five of the 26 patients (96.2%) with diabetes were seropositive for anti-GAD antibodies. Serum anti-GlyR antibody testing was documented for 17/69 (24.6%), and 11/69 (15.9%) were positive. Eight of these 11 had PERM. Testing for other antibodies was documented in 26/69 (37.7%), and seven were positive for the following antibodies: three for anti-thyroid antibodies; two against parietal cells; one for anti-amphiphysin; one for rheumatologic antibodies (Anti-Sjögren's-syndrome-related antigen A and anti-nuclear antibody);²³ one for non-specific immunoglobulin (Ig)M against tick-borne meningoencephalitis; and one for anti-N-methyl-D-aspartate receptor antibodies. Four of these seven patients had two or more detectable autoantibodies in serum.^23–26 Only the patient with antibodies against tick-borne meningoencephalitis had them in both serum and CSF.⁶ Testing for oligoclonal bands (OCBs) was reported in 17 patients (24.6%), and they were detected in nine. Electromyography (EMG) findings were reported in 36 patients (52.2%), with findings consistent with SPS in 26.

As shown in Table 2, the two most commonly used treatments were benzodiazepines and intravenous immunoglobulin, which were used in 45/69 cases (65.2%). Baclofen followed as the third most commonly used agent (26/69 cases, or 37.7%). Fifty-eight patients (84.1%) received either intravenous immunoglobulins, plasmapheresis, rituximab, steroids, or a steroid-sparing immunosuppressant. In terms of outcomes, 54 reported improvement (78.3%), nine (13%) remained stable, and three (4.3%) either worsened or died (one of these had PERM; the other two had SPS, one of which was also had concomitant corticobasal syndrome).^27–29 Eighteen of the 54 (33.3%) patients who improved experienced at least one relapse at some point in the course of their disease.

Paraneoplastic SPS-spectrum cases (Table 3)

Table 3. Reported Cases of Paraneoplastic Stiff Person Syndrome and Its Variants.

				Symptoms						History or Examination findings													Antibody Testing					Other Testing			Treatment													Outcome
Reference	Reported Diagnoses	Age And Sex At Presentation	Number Of Symptomatic Months Before Presentation	Bulbar Dysfunction	Falls	Gait Difficulties (Including Ataxia)	Pain	Axial Stiffness Or Hyperlordosis	Autonomic Instability	Concomitant Autoimmunity (Other Than Diabetes)	Cramps/Spasms	Diabetes	Extraocular Movement Abnormalities	Encephalopathy Or Psychiatric/Behavioral Changes	Hyperekplexia	Hyperreflexia	Limb Posturing/Dystonia	Limb Stiffness/Rigidity	Malignancy	Myoclonus	Seizures	Weakness	Serum Anti-GAD Antibodies	Serum Anti-Glyr Antibodies	Other Serum Antibodies Detected	Other CSF Antibodies Detected	CSF Oligoclonal Bands	Electroencephalographic Abnormalities	Electromyography Findings Consistent With SP	MRI Brain Abnormalities (Incidental Findings Excluded)	Antibiotics	Antiepileptics	Baclofen	Benzodiazepines	Botulinum Toxin Injections	Chemotherapy Or Radiotherapy	Intravenous Immune Globulin	Other Agents Or Interventions	Plasmapheresis	Rituximab	Steroids	Steroid-Sparing Immunosuppression	Surgical Intervention	Improvement Or Resolution	Stabilization Without Improvement	1 Or More Relapses Mentioned	Worsening And/Or Death
Agarwal et al.75	SL	55F	2			+	+				+				+	–	+	+	B				+		–		–	–	+	–			+	+		+					+		+	+		+
Kosseifi et al.76	SP	29M	<1	+		+	+				+							+	T			+	–		–									+		+		+						+
Schmidt et al.77	PERM	21M	18		+	+			+				+	+	+			+	H	+	+		–		–		–	–	+	–	+	+		+		+		+			+			+
Thumen et al.78	OM,SP	66F	36	+	+	+		+					+	+	+	+	+	+	B				–	–	AR	AR	+		+	–			+	+							+		+	+
Chamard et al.79	SL	65F	<1			+	+		+		+	+			+		+	+	B						AA	AA	+		+	–			+	+		+		+			+		+	+		+
Lemieux et al.80	PERM	30F		+	+	+	+				+		+	–		+		+	B		–	–	–		–			–	–	–			+	+		+	+	+	+				+	+		+
Byrne et al.81	LE,SP	60F	4		+	+	+		+		+			+		+	+	+	B			+	–		AA		+	+		+			+	+		+	+	+		+	+		+	+		+
Krishna et al.82	LE,SP	54F	2				+	–			+		–	+	+	–		+	B		+	–	–		AA		+			+		+	+	+		+								+
Aghajanzadeh et al.83	SP	32M						+			+	–						+	T		–		+										+	+				+					+	+
Badzek et al. 84	SP	55F				+		+										+	C				–		–									+		+							+	+
Derksen et al.38	SL	66M	1			+	+		+	–	+	–			+	+	+	+	K	+			–	+	–		–		+	–			+	+		+	+		+	+					+
Rakcevic et al.85	SP	57F				+	+	+			+				+				L				+		–								+	+		+			+	+	+	+		+
Kelly et al.86	SP	64F			+		+		–		+			+	+			+	B						AA	AA							+		+	+	+	+	+	+	+		+	+
Kobayashi et al.87	SP	81F	156	+		+	+	+			+	–			+				T				+																		+		+	+
Koca et al.88	SP	58F		+		+	+	+			+	–	–		+	–		+	ME		–	–	+				–		+	–			+	+		+		+			+			+			+
Laroumagne et al.31	OM	65M				+							+						L	+			+		–			–		–						+					+						+
Pagano et al.30	WNE	72M																	C				+														+		+			+			+
	PN	54F		+		+													B				–														-		+			-		+
	A,SP	70F		+		+													N				+														+		+			-		+

Abbreviations: A, ataxia; AA, anti-amphiphysin antibodies; AR, anti-Ri antibodies; B, breast cancer; C, colon cancer; CSF, cerebrospinal fluid; F, female; H, Hodgkin lymphoma; K, leukemia; L, lung cancer; LE, limbic encephalitis; M, male; ME, mesothelioma; MRI, magnetic resonance imaging; N, melanoma; OM, opsoclonus-myoclonus; PERM, progressive encephalomyelitis with rigidity and myoclonus; PN, paraneoplastic centrally mediated disorder with central planning impairment; SL, stiff limb syndrome; SP, stiff person syndrome; T, thymic malignancy; V, vitiligo; WNE, west Nile encephalitis

There were 19 paraneoplastic SPS-spectrum cases. Thirteen were females. The median age at presentation was 59 (range 21–81) years. Eleven had SPS, four of whom also had other neurologic disorders (two with limbic encephalitis; one with ataxia; and one with opsoclonus–myoclonus). Three cases were consistent with SLS and two with PERM. One had West Nile encephalitis with positive anti-GAD antibodies; another had a “paraneoplastic centrally mediated disorder with central planning” but had no detectable antibody; and the remaining case had isolated opsoclonus–myoclonus with serum anti-GAD antibodies.^30,31 Ten cases documented the time between symptom onset and presentation with a median of 3 months and ranging from weeks up to 13 years. Fifteen patients described gait difficulties, 11 reported pain, 12 had cramps or spasms, and 13 had limb stiffness.

All cases had a documented malignancy, the most common being breast cancer (eight cases). Malignancies of the following organs were also reported: thymus (three cases); colon (two); and lung (two). The remaining four cases had either Hodgkin’s lymphoma, leukemia, mesothelioma, or melanoma. In terms of comorbidities, only one had pre-existent diabetes (another four were documented not to be diabetic) and screening for autoimmunity was documented in only one case (and not detected).

The antibody profile was varied. Seventeen cases were tested for anti-GAD antibodies, of which eight were positive. Anti-GlyR antibodies were tested in two cases, one was positive. Thirteen cases were tested for other antibodies, and the following were identified: four were seropositive for anti-amphiphysin antibodies (two cases had antibodies also in CSF) and one was positive for anti-Ri antibodies in both serum and CSF. No antibodies were detected in five. Oligoclonal bands were tested in eight, and four were positive. Seven had EMGs, with six having features consistent with SPS.

Benzodiazepines and cancer treatment (chemo- or radiotherapy) were employed in 13 cases. Nine cases underwent oncologic surgery, and 16 patients received some kind of cancer treatment, including chemotherapy, radiotherapy, and/or oncologic surgery. Despite all of these patients having a malignancy, 16 improved neurologically (one of them deteriorated eventually). Two remained stable and two others worsened, one of whom succumbed to mesothelioma after having previously improved from a neurologic standpoint. Four of the 16 patients who improved experienced at least one relapse.

Cryptogenic SPS-spectrum cases (Table 4)

There were 13 cryptogenic cases. Four were female, nine were male. The ages ranged from 7 to 75 years, with a median age of 44 years. Of these 13, 11 had SPS, including a patient who initially presented with SLS.³² Among the two remaining cases, one had SLS and the other had PERM. Data for 11/13 patients regarding the time between symptom onset and presentation ranged from less than 1 month to 14 years, with the median being 9 months. All of the cases had either cramps or spasms, and nine had limb stiffness. Other common symptoms were pain (reported in 11 cases) and gait difficulties (nine). Only one case had concomitant autoimmunity (thyroid disease and vitiligo),³³ and absence of concomitant autoimmunity was only reported in another case.¹⁴ Eleven of 13 reported testing for commonly associated antibodies, and all of them were negative for either anti-GAD or anti-GlyR antibodies. Of these 11, one reported positive OCBs³⁴ while another documented absence of this finding (the patient with PERM).³⁵ EMG findings consistent with SPS were reported in seven. The most commonly used medication was a benzodiazepine (12 cases). Treatment outcomes in this cohort were overwhelmingly positive. Twelve reported improvement and the thirteenth stabilized, but there was no mention of worsening or death. Among the 12 who improved, only the PERM case relapsed.³⁵

Case series of patients with anti-GAD antibodies and cerebellar ataxia (Table 5)

Table 5. Case Series of Patients with Anti-GAD Antibodies in the Setting of Cerebellar Ataxia with or without Stiff Person Symptoms.

				Number of Patients Reporting the Following Symptoms		Number of Patients with the Following history of Examination Findings							Number of Patients with the Following Work-up Findings			Number of Patients Treated with the Following Medications			Number of Patients with the Following Outcome to Treatment
Reference	Reported Diagnoses at Presentation (n)	Age Range and Sex Distribution	Number of Symptomatic Months Before Presentation (n)	Bulbar	Gait Difficulties (Including Ataxia)	Concomitant Autoimmunity (Except Diabetes) (N)	Cramps/Spasms	Diabetes	Extraocular Movement Abnormalities	Hyperekplexia	Malignancy	Seizures	Serum Anti-GAD Antibodies	CSF Oligoclonal Bands	Electromyography Findings Consistent With SP	Intravenous Immune Globulin	Rituximab	Steroids	Improvement or Resolution	Stabilization Without Improvement	Worsening and/or Death
Arino et al.21	A (6) A,SP (28)	33–80 28F, 6M	<1 (13)	24	31	T (18) P (7) V (2)	9	13	20	1	4	4	34	16	4	10	1	10	10	5	3

Abbreviations: A, Ataxia; CSF, Cerebrospinal Fluid; F, Female; GAD, glutamic acid decarboxylase; M, Male; MRI, Magnetic Resonance Imaging; n, Number of Cases; P, Pernicious Anemia; T, Autoimmune Thyroid Disease; SP, Stiff Person Syndrome; V, Vitiligo.

Ariño et al.²¹ described 34 cases of cerebellar ataxia with anti-GAD antibodies. Data are consolidated for the cohort, as individual details for each of these 34 are not available. Twenty-eight of the 34 patients were female. Gait difficulty was the most common symptom in this cohort (91.2%). Concomitant autoimmunity was reported for well over half, with thyroid disease being the most prevalent (52.9%). Cancer was detected in four cases (11.8%). Different immunosuppressant agents were the treatment of choice, and outcomes varied considerably with a subacute course and early medical treatment being good prognostic indicators.

Discussion

Segregation of SPS-spectrum cases based on likely etiology allowed detecting similarities and differences between categories. It also allowed identifying themes within each group. Perhaps the most evident observation is the phenotypical variability present within each group. The data demonstrate that SPS, SLS, or PERM can all be found in autoimmune, paraneoplastic, or cryptogenic SPS-spectrum disease. This lack of specificity hinders the examiner’s ability to predict an etiology based solely on clinical features.

By definition, the presence of anti-GAD antibodies includes a condition within the SPS spectrum. However, these antibodies do not differentiate between paraneoplastic and autoimmune causes as they are present in both. In this review over 80% of autoimmune SPS-spectrum cases were positive for anti-GAD antibodies, but close to 50% of paraneoplastic cases were also seropositive. Additional antibodies (anti-amphiphysin antibodies) can increase specificity, particularly when multiple antibodies are present, in which case anti-amphiphysin antibody seropositivity should encourage an investigation for an underlying malignancy. The presence of multiple antibodies further supports the notion that anti-GAD antibodies are not the sole pathogenic cause but that the GABA synthesis apparatus along with its associated pathways are disrupted. It also suggests that it is important to obtain CSF in these patients with multiple antibodies as some are predominantly found in the CNS, such as GAD67,²¹ particularly when clinical findings or preliminary serum tests are inconclusive.

Anti-GlyR antibodies were the second-most commonly observed after anti-GAD antibodies. In this review, 11 autoimmune SPS-spectrum cases had anti-GlyR antibodies in serum (15.9%), nine had them in CSF (13%), seven had antibodies in both CSF and serum (10.1%), and two only in CSF but not in serum (2.9%). The phenotypes included PERM, three with SPS, one with stiff trunk syndrome, and one with behavioral changes and epilepsy. Interestingly, among the patients who had antibodies in both CSF and serum, one had prodromal reduction in taste,²⁵ one had prodromal pruritus,³⁶ and a last one had both.³⁷ All of them were eventually diagnosed with PERM. We find this observation clinically relevant, since changes in taste and pruritus may be specific enough to predict a diagnosis of PERM in the right clinical context. This could ultimately translate into prompt initiation of immunosuppressive treatment. In contrast to autoimmune cases, anti-GlyR antibodies were only found in one paraneoplastic case: a case of SLS in the setting of leukemia.³⁸ Thus, the presence of anti-GlyR antibodies despite the clinical phenotypic variation may strongly suggest a more autoimmune pathology rather than a paraneoplastic variation, unlike anti-GAD antibodies, which are present in both.

Paraneoplastic SPS-spectrum cases too present with a myriad of different phenotypes. Because of this, it may be prudent to screen for malignancy in all patients with symptoms consistent with the SPS-spectrum, although, admittedly, it may not be cost-effective. Thus, we recommend a step-wise approach, first with brain and spine MRI, routine blood work, including complete blood count, comprehensive metabolic panel, and B12 level, EMG/nerve conduction studies, and lumbar puncture assessing for autoimmune and infectious causes before a malignancy screen. If this preliminary work-up and subsequent serum and CSF paraneoplastic antibody panels are negative, whole-body positron emission tomography is recommended to assess for occult malignancy. The possible scenario in which this intervention may be most cost-effective is in patients with opsoclonus–myoclonus. We identified the opsoclonus–myoclonus phenotype to be specific for paraneoplastic cases, as seen in two out of 19 paraneoplastic cases, as opposed to none of the 69 autoimmune or the 13 cryptogenic cases. However, if serum anti-amphiphysin antibodies are detected, a specific search for breast malignancy should be undertaken as previous reports^19,39 and our current analysis show a high likelihood of associated breast cancer.

Among ancillary testing, EMG is particularly useful. Although not consistently reported, EMG was positive in 72.2% of all EMG-tested autoimmune cases, 85.7% of paraneoplastic cases, and 87.5% of the cryptogenic EMG-tested cases. Given this apparent high probability of a confirmatory finding, we advocate for routine EMG testing in all suspected SPS-spectrum cases featuring rigidity, dystonic posturing, or cramping. However, there were no specific EMG findings to differentiate among the subgroups. One important consideration based on both our literature review and experience is that the pathognomonic EMG findings may take some time to develop.³³ Thus, in cases where an initial EMG is negative, a repeat EMG should be performed in at least 3 months’ time if the clinical suspicion is high.

Thus, the constellation of commonalities within groups is most helpful when attempting to identify the cause of a SPS-spectrum condition, rather than relying solely on clinical features or a single antibody. Thus, the presence of a concomitant autoimmune condition, anti-GAD antibodies, and symptoms of a SPS-spectrum disorder suggests an autoimmune pathomechanism. However, anti-amphiphysin antibodies in the presence of these symptoms and a malignancy suggest a paraneoplastic etiology. Although this may not be absolute, it can provide a guide to quickly diagnose and appropriately treat these patients. For example, in patients with paraneoplastic disease, treatment of the underlying malignancy with chemo/radiotherapy and oncologic surgery should be the focus of care.

Whereas we are confident of our autoimmune and paraneoplastic classifications, we have reservations with the cryptogenic category. We believe that this category can potentially include autoimmune cases for which a known antibody is either present in titers that evade detection, or are positive for antibodies that are yet to be recognized. The response to immunosuppression (plasmapheresis, steroids) in five out of 13 cases argues in favor of this theory. Of note, these patients also received other interventions, and it is impossible to ascertain which agent(s) was responsible for the symptomatic response. In addition, the presence of a microscopic malignancy in some of these cases cannot be completely excluded, and patients with cryptogenic SPS-spectrum disease may eventually declare themselves as paraneoplastic, granted a sufficient follow-up period is allowed. Finally, the absence of both antibodies and EMG evidence to support a diagnosis of SPS-spectrum disease opens the possibility for cramping syndromes or other conditions featuring rigidity to be misclassified within the SPS spectrum. The differential diagnosis of SPS is vast and includes myelopathy, myopathy, Isaac’s syndrome, Parkinson’s disease and atypical parkinsonian syndromes, primary lateral sclerosis, ankylosing spondylitis, neuroleptic malignant syndrome, serotonin syndrome, hereditary or tropical spastic paraparesis, spinal interneuronitis with rigidity, dystonia, neuromyotonia, and tetanus.^6,37 With this differential in mind, in seronegative patients with inconclusive EMGs, a comprehensive work-up can include a complete blood cell count, a comprehensive metabolic panel, thyroid function tests, creatine kinase, erythrocyte sedimentation rate, C-reactive protein, serum B12 level, human T-cell lymphotropic virus-1, rheumatoid factor, antinuclear antibody, computed tomography (CT) chest screening for a mediastinal mass, CT abdomen and pelvis, MRI brain, and CSF analysis including oligoclonal bands and IgG index.¹⁵

With these caveats in mind, the great majority of cryptogenic cases were diagnosed with SPS. Interestingly, unlike autoimmune and paraneoplastic cases, the majority of cryptogenic cases were male. This may be an indication of a different pathophysiology within this group, but the significance of this finding is unclear. The majority of cryptogenic cases in this review reported symptomatic improvement. However, some of the success was with novel interventions for SPS (including spinal cord stimulation,³² physical therapy,⁴⁰ dantrolene,⁴¹ intrathecal baclofen,⁴² and cannabis⁴³). We find this likely constitutes a reporting bias, since it is possible that cases in which novel interventions were unsuccessful may not have been published. Thus, we wonder whether the prognosis of these patients is as favorable as the data in this review suggested.

There are certain limitations to consider regarding our data. To start, the prevalence estimates of SPS, PERM, and other variants among the three subdivisions were based on the original published descriptions. It was difficult for us to ensure that the specific criteria established by Dalakas were met, as clinical descriptions were often incomplete. We are thus relying on the original author’s clinical judgment in diagnosing these patients with the said conditions. It is possible that those diagnosed with classical SPS may in fact have been a variant and vice versa. In addition, both the paraneoplastic and the cryptogenic groups had fewer than 20 cases each, making inferences from these groups difficult to generalize. Furthermore, many of the cases reported describe treatment successes, and it is possible that this overestimates the response rates of patients with these conditions. In addition, it is impossible to demonstrate the complete absence of cancer in all of the autoimmune cases, just as it is impossible to guarantee that an autoimmune (and not a paraneoplastic) etiology is not responsible for SPS symptoms in some of the paraneoplastic cases. Finally, it is unclear whether the data acquired would vary significantly should our search be expanded to include cases in other languages, or cases reported before 2010.

Overall, this review supports the idea that, unlike the original descriptions, SPS encompasses a spectrum of diseases related by their clinical symptoms, the presence of autoantibodies, EMG findings, and its response to immunomodulation and muscle relaxants. Despite these commonalities, different causative mechanisms are likely among the disorders, and their classification based on likely etiology can guide treatment and provide useful prognostic information. As new associated antibodies and associated clinical features are discovered, it is possible that the spectrum will continue to expand. Ultimately, a high degree of clinical suspicion is necessary to initiate the work-up and select appropriate treatment that can be initiated in a timely fashion. Future directions in the management of these conditions may emerge as we better understand the role, if any, that these antibodies play in the emergence of symptoms. This could lead to the development of targeted therapies that minimize systemic side effects.