Morvan syndrome and related disorders associated with CASPR2 antibodies
Jan. 23, 2023
At vero eos et accusamus et iusto odio dignissimos ducimus qui blanditiis praesentium voluptatum deleniti atque corrupti quos dolores et quas.
Motor neuropathies and multifocal motor neuropathy with conduction block are treatable causes of neuropathy that present with the clinical syndrome of asymmetric motor weakness with atrophy, which may mimic amyotrophic lateral sclerosis. The diagnosis of the disorder relies on the typical clinical presentation as well as electrodiagnostic findings of conduction block at noncompression sites and often on anti-GM1 IgM antibodies on serum testing. The mainstay of therapy includes intravenous immunoglobulin, but often other immunomodulating therapy is required for long-term management. In this article, the author summarizes the most recent literature covering pathophysiology, diagnostic criteria, and treatment for motor neuropathy.
• Multifocal motor neuropathy is a treatable cause of neuropathy that may mimic amyotrophic lateral sclerosis.
• The diagnosis of multifocal motor neuropathy relies on the typical clinical presentation, electrophysiology demonstrating conduction block outside of compression sites, and the presence of anti-GM1 IgM antibodies.
• The primary diseases to consider in the differential diagnosis of multifocal motor neuropathy include amyotrophic lateral sclerosis, chronic inflammatory demyelinating polyneuropathy, and Lewis Sumner syndrome (multifocal acquired demyelinating sensory and motor neuropathy).
• Motor neuropathies and multifocal motor neuropathy often respond to treatment with intravenous immunoglobulin, but in some patients, long-term management requires the use of other immunosuppressant medications.
Motor neuropathy is manifested by weakness, often with muscle atrophy and fasciculations and normal sensory functions. It can be generalized, or involve 1 or several nerves, and can be slowly progressive or acute in onset. The motor axons or nerve fibers are primarily affected, distinguishing it from motor neuron disease wherein the perikaryon or nerve cell body is primarily involved. Motor neuropathy is often immunologically mediated, and can respond to immunosuppressive therapies, whereas motor neuron disease does not.
The majority of clinical research, epidemiology, and treatment trials assessing motor neuropathy have focused on multifocal motor neuropathy, also known as multifocal motor neuropathy with conduction block. This describes the clinical entity of asymmetric motor neuropathy, often with conduction block or other demyelinating findings on electrodiagnostic studies and sometimes with anti-GM1 ganglioside antibodies. Less typical clinical pictures are also included in the framework of multifocal motor neuropathy, including a more symmetric neuropathy, those without conduction block, and those without GM1 antibodies; these neuropathies are at times referred to as “motor neuropathy” without using the term multifocal motor neuropathy but in many series and clinical care are considered a shared entity in terms of pathophysiology and management. A separate category is motor neuropathies associated with other causes, including those with amyloidosis. These motor neuropathies should be clinically managed according to their associated etiology and generally considered separately from the spectrum of multifocal motor neuropathy. Except where specifically noted, the majority of this article addresses the entity of what is commonly referred to as multifocal motor neuropathy when describing motor neuropathies.
Motor neuropathy is described in papers dating back to the 1920s (144; 70; 185). Interest in the syndrome, however, has flourished since the 1980s with the discovery of autoantibodies to glycolipid neural antigens and nerve conduction abnormalities, such as conduction block, which have aided in identifying patients with motor neuropathy.
Peters and Clatanoff described a patient with spinal muscular atrophy and an IgM monoclonal gammopathy who improved following treatment with chlorambucil (138). In retrospect, that patient probably suffered from motor neuropathy, rather than from motor neuron disease. The first documented case of a patient with motor neuropathy and an IgM monoclonal gammopathy was reported by Rowland and colleagues (146). That patient presented with progressive weakness, muscle atrophy and fasciculations, and had an IgM kappa monoclonal gammopathy. Motor conduction velocities were slow, with normal sensory conductions. Postmortem examination revealed a normal number of anterior horn cells with central chromatolysis (implicating a peripheral neuropathy) and severe loss of nerve fibers in the anterior roots.
Multifocal demyelinating neuropathy with persistent conduction block was first described by Lewis and colleagues. They reported several patients with a chronic demyelinating sensorimotor neuropathy, characterized clinically by mononeuritis multiplex and electrophysiologically by persistent multifocal conduction block (103). In 1985, Parry and Clarke described several patients with a syndrome resembling motor neuron disease, but with multifocal motor conduction block and normal sensory conductions (133).
IgM anti-GM1 antibodies were first described by Freddo and associates in a patient with lower motor neuron syndromes and IgM monoclonal gammopathy (57). A second patient with the same syndrome improved with immunosuppressive therapy (98). Pestronk and colleagues reported that patients with multifocal motor neuropathy and conduction block also had high titers of IgM anti-GM1 antibodies and coined the term “multifocal motor neuropathy” (136).
Katz and colleagues have proposed a separate entity known as “multifocal acquired motor axonopathy” (or MAMA neuropathy) to describe those patients with pure motor neuropathy without electrophysiologic conduction block or demyelination. Some of these patients have GM1 antibodies, and clinically they appear to respond similarly to therapy (81; 56). Although the authors of these reports argue that there are differences in the clinical and serologic qualities of the patients described, many feel that this subset of patients likely represents the same pathophysiologic spectrum as multifocal motor neuropathy, but in which the typical electrophysiologic findings of conduction bock cannot be captured anatomically or temporally.
Motor neuropathy, and particularly multifocal motor neuropathy, is characterized by weakness with muscle wasting and fasciculations, and may resemble motor neuron disease. It can begin at any age, with onset reported between the ages of 15 and 79 years. Men are more frequently affected than are women. It is classically progressive or worsens in a stepwise fashion and is often asymmetric, affecting distal muscles, more often in the arms than the legs. Sensory symptoms are notably absent, with only the occasional complaint of paresthesias. Motor weakness occurs typically in the distribution of individual motor nerves rather than the segmental distribution seen in amyotrophic lateral sclerosis (95; 08; 159; 118; 165; 153). In the arms, muscle groups most often affected include finger extensors or finger flexors (153). Many patients with multifocal motor neuropathy exhibit striking differences in the weakness of the muscles innervated by the same motor nerve, particularly in the finger extensors (28; 72).
When there is prominent conduction block, significantly weak muscles may have relatively normal bulk. Uncommonly, true hypertrophy of proximal muscles, attributed to constant activation by ectopic motor action potentials in the corresponding nerve, has been described (127; 21). Although bulbar involvement or respiratory failure can occur, they are rare, unlike motor neuron disease (79; 107; 17). Deep tendon reflexes may be absent or reduced but are sometimes active in weak and wasted limbs (77; 84; 50), although Babinski signs are invariably absent and spasticity does not occur. Myokymia rarely occurs (145; 20).
Compared with amyotrophic lateral sclerosis, motor neuropathies have a favorable prognosis and a prolonged course. Although multifocal motor neuropathy with conduction block has only recently been recognized, a few patients have died after symptoms of 20 years. There is a beneficial effect of therapy on the natural history of multifocal motor neuropathy: a longer number of years without immunoglobulin therapy has been identified as an independent determinant of disability and permanent weakness in patients with multifocal motor neuropathy (28). Whereas the majority of patients with amyotrophic lateral sclerosis typically die from respiratory failure within 2 to 5 years, respiratory failure with multifocal motor neuropathy and conduction block has only rarely been described but may occur after a prolonged course (95; 107; 131). Life expectancy is normal for patients with multifocal motor neuropathy despite its unremitting course.
At the age of 48 years, a woman developed mild paresthesias of her fingers bilaterally. At the age of 52 years, she developed a right-hand tremor that was followed by weakness. The tremor moved to the left hand. At the age of 53 years, she noted bilateral toe numbness and slowness of gait. She had no bulbar symptoms.
On neurologic exam, she had normal mental status and cranial nerve function. She had asymmetric weakness in the left arm greater than the right arm. She had 4/5 strength in the deltoids, biceps, and triceps bilaterally, finger and wrist extensors were 1/5, interossei and abductor pollicis brevis strength was 3/5, and finger flexors were 4/5. There was no atrophy or fasciculations. In the lower extremities, tibialis anterior were 4/5, and other muscles in the leg were 5/5. Sensation was normal. Deep tendon reflexes were absent. There was no Babinski sign.
Three EMGs were performed over 3 years. They revealed low motor evoked responses in the median, ulnar, peroneal, and tibial nerves with normal conduction velocity and normal or only minimally prolonged distal motor latencies and F wave latencies. There was no conduction block or abnormal temporal dispersion. Sensory evoked response amplitudes were normal. Creatine kinase was minimally elevated at 304 (normal is less than 295). The following tests were normal; complete blood count, erythrocyte sedimentation rate, chemistry panel, and thyroid-stimulating hormone. Serum protein electrophoresis was initially normal but revealed a monoclonal IgM lambda on later testing. Anti-GM1 antibodies were positive at 1 to 51,000 (normal is less than 800) by ELISA. Cerebrospinal fluid showed 3 white cells and a protein of 40. IgG total protein was 5.8% (normal < 13%). Bone marrow biopsy was normal. A muscle and nerve biopsy of the motor and sensory nerve to the gracilis muscle was performed. The muscle biopsy revealed neurogenic atrophy with fiber type grouping. The sensory nerve biopsy was normal. The motor nerve biopsy revealed a loss of large diameter myelinated fibers, thinly myelinated fibers, and regenerative clusters.
The patient was initially treated with high-dose intravenous immune globulin (2 g/kg); it was effective for over 1.5 years but became ineffective. Plasmapheresis did not help. She was given prednisone 60 mg and diffusely worsened, including weakness. She was given rituximab, and the symptoms began to stabilize. Treatment with intravenous immune globulin was restarted with improvement in her weakness.
Comment. This patient's history illustrates several points. She developed progressive asymmetric weakness. She had minimal sensory symptoms, although she had no objective sensory findings on clinical examination or on nerve conduction studies. Her motor nerve conduction studies were abnormal but did not show conduction block or other signs of demyelination. She had markedly elevated antibody titers to GM1. She responded initially to intravenous immune globulin, but after 1 year she no longer responded. Her response to intravenous immune globulin was restored by treatment with rituximab.
Though the etiology of motor neuropathies is uncertain, many motor neuropathies are thought to be immune-mediated because of the presence of the autoantibodies and the response to immunomodulating therapy.
The evidence that autoantibodies, such as anti-GM1 antibodies, are relevant to the pathogenesis of motor neuropathies is discussed in the next section.
The cause of weakness in motor neuropathies, and multifocal motor neuropathy in particular, is likely related to a combination of demyelination at areas of conduction block and a component of focal and generalized axonal loss. In 1 study of 20 patients with multifocal motor neuropathy on IVIG, EMG evidence of axonal loss was common, occurring in 61% of muscles tested, findings of axonal loss were the most significant determinant of muscle weakness, and conduction block was an independent determinant of axonal loss (164).
Pathological studies of motor neuropathy are rare. In 1 study, nerves at the site of conduction block in 2 patients with anti-GM1 antibodies showed evidence for demyelination without inflammatory cell infiltrates (78). Similar changes were seen in other patients with motor neuropathy and conduction block, although not clearly at the site of conduction block (07; 37). Another study of 7 patients with pathological specimens at the site of motor conduction block showed no signs of demyelination, but instead showed multifocal axonal degeneration (158). IgM deposits at the nodes of Ranvier have been seen in patients with multifocal motor conduction block and anti-GM1 antibodies (148; 02).
Nerve biopsy can be helpful in making a diagnosis of motor neuropathy. Motor nerve biopsies of the obturator nerve to the gracilis muscle from patients with motor neuropathy showed an increase of regenerative clusters of small myelinated fibers, in comparison to those in patients with motor neuron disease (37). Sural nerve biopsies show no abnormalities (172; 65) or show only minimal abnormalities (65; 42).
A few patients with motor neuropathy have been examined at postmortem. A patient with motor neuropathy and an IgM monoclonal gammopathy was reported by Rowland and colleagues (146). Postmortem examination revealed a normal number of anterior horn cells, with central chromatolysis and severe loss of nerve fibers in the anterior roots. Another patient with motor neuropathy and an IgM polyclonal gammopathy at postmortem had a loss of anterior horn cells, also with central chromatolysis and a loss of anterior root fibers (134). Pathological studies at postmortem in a patient who died with motor neuropathy and elevated titers of anti-GM1 antibodies revealed that the patient also had degeneration of the anterior roots, immunoglobulin deposits on myelin sheaths, and chromatolytic changes in spinal motor neurons (02). The predominant involvement of the anterior roots, rather than the more distal nerve segments, might explain the lack of correlation between the presence of conduction block and the distribution and severity of the weakness in many of the affected patients. The motor neurons may be secondarily involved to the anterior roots, as suggested by the presence of central chromatolysis.
Motor neuropathy and monoclonal gammopathy. Motor neuropathy is sometimes associated with monoclonal gammopathy resulting from the abnormal proliferation of monoclonal B-cells that secrete excessive IgM, IgG, or IgA antibodies (97). The monoclonal antibodies are detected and characterized by serum protein immunoelectrophoresis or immunofixation electrophoresis and are also called M-proteins or paraproteins. M-proteins can sometimes be autoreactive and cause autoimmune disease, but approximately 1% of normal adults have serum M-proteins, so that the monoclonal gammopathy can be coincidental and unrelated to the associated disease.
In patients with neuropathy and IgM M-proteins, the monoclonal gammopathy is usually nonmalignant, although it is sometimes associated with Waldenstrom macroglobulinemia, B-cell lymphoma, or chronic lymphocytic leukemia. The incidence of peripheral neuropathy in patients with IgM monoclonal gammopathies has been reported to be between 5% and 50%. In most cases, the IgM M-proteins have autoantibody activity and react with oligosaccharide determinants of glycolipids or glycoproteins (glycoconjugates) concentrated in peripheral nerve.
Structure of glycolipids. Glycosphingolipids are composed of the long-chain aliphatic amine sphingosine (acylated ceramide) attached to 1 or more sugars. Gangliosides are complex glycosphingolipids containing sialic acid. Sialic acid is a generic term for N-acylneuraminic acid. Gangliosides are designated G for ganglioside followed by M, D, T, or Q (for mono, di, tri, or quad) referring to the number of sialic acids. Arabic numbers and lowercase letters follow and refer to the sequence of migration by thin layer chromatography (155; 183).
Motor neuropathy and anti-GM1 antibodies. In most patients, the anti-GM1 antibodies recognize the Gal(B1-3)GalNAc determinant that is shared by asialo GM1 (AGM1) and the ganglioside GD1b. The same determinant is also present on some glycoproteins and is recognized by the lectin peanut agglutinin. Some of the antibodies, however, are highly specific for GM1, or recognize internal determinants shared by GM2 (71; 10; 91; 147). Although GM1 and other Gal(B1-3)GalNAc-bearing glycoconjugates are highly concentrated and widely distributed in the central and peripheral nervous systems, they are mostly cryptic and unavailable to the antibodies. However, anti-GM1 antibodies bind to spinal cord grey matter, and to GM1 on the surface of isolated bovine spinal motor neurons, but not to dorsal root ganglion neurons (39). In peripheral nerve, GM1 ganglioside and Gal(B1-3)GalNA-bearing glycoproteins are expressed at the nodes of Ranvier (38; 151). Two of the glycoproteins have been identified as the oligodendroglial-myelin glycoprotein in paranodal myelin, and a versican-like glycoprotein in the nodal gap (05). The antibodies also bind to the presynaptic terminals at the motor endplate in skeletal muscle, where the antibodies might also exert an effect (98; 160).
Although anti-GM1 IgM antibodies are generally thought to be pathogenic in the development of motor neuropathies and multifocal motor neuropathy, it is not clearly known whether they definitively cause disease or are an associated abnormality. The binding to motor neurons, but not to sensory neurons, correlates with the clinical syndrome, and GM1 is highly enriched in myelin sheaths of motor nerves and differs in its ceramides in comparison to sensory nerves (122; 123). This might render the anterior roots more susceptible to the autoantibodies' effects. In 1 study, rabbits immunized with GM1 or Gal(1-3)GalNAc-BSA developed conduction abnormalities with immunoglobulin deposits at the nodes of Ranvier (161), and in another study, serum from a patient with increased titers of anti-GM1 antibodies and IgM deposits at the nodes of Ranvier produced demyelination and conduction block when injected into rat sciatic nerve (149). The human anti-GM1 antibodies have also been shown to bind to and kill mammalian spinal motor neurons in culture (66) and at the motor endplate (183). Serum from patients with multifocal motor neuropathy, both with and without anti-GM1 antibodies, can block nerve conduction in the mouse phrenic nerve-diaphragm preparation (143). Conflicting data show the ability of anti-GM1 antibodies to alter potassium current and, in the presence of complement, to block sodium channels in rat myelinated nerve fibers (156; 14; 75; 181). Based on the persistence of the motor conduction block, and on the pathological findings of axons devoid of myelin and only minimal onion bulbs, Kaji and colleagues suggest that anti-GM1 antibodies impair remyelination (75). The variable regions of anti-GM1 antibodies from normal individuals or patients with neuropathy, exhibit multiple somatic mutations in their hypervariable regions, suggesting that they have been derived from a process of antigenic stimulation (182; 109).
In contrast to the IgM anti-GM1 antibodies in patients with chronic motor neuropathies, increased titers of polyclonal IgG or IgA anti-GM1 antibodies are associated with an acute motor axonal neuropathy, which is a variant of the Guillain-Barré syndrome. These antibodies have been reported to occur following infection with Campylobacter jejuni (190; 180; 171; 114; 86), which bears GM1-like oligosaccharides (06; 188; 186), or following parenteral injection of GM1 containing gangliosides (99; 120; 94). Postmortem studies in some patients who died of the Guillain-Barré syndrome following Campylobacter jejuni infection, show noninflammatory degeneration of the anterior roots, and chromatolytic changes in spinal motor neurons (114), similar to the chronic disease associated with IgM anti-GM1 antibodies. The Penner O to 19 serotype is overrepresented in patients who develop Guillain-Barré syndrome, compared with those who only have enteritis following Campylobacter jejuni infection. However, some evidence suggests that host factors play a role in developing Guillain-Barré syndrome (51).
Anti-GD1 IgM autoantibodies. Several patients with motor neuropathy and anti-GD1a antibodies have been described. The initial description was of a 73-year-old man with 3 years of lower limb weakness and an IgM k monoclonal gammopathy. He had absent leg and reduced arm reflexes. Cerebrospinal fluid exam was normal and nerve conduction velocity was slow. During treatment with melphalan and corticosteroids, there was no progression for 2 years (18). There have been other limited reports of motor neuropathies associated with GD1a antibodies variably responding to immunotherapy (26; 76). In 1 series of patients with multifocal motor neuropathy, 9% of patients had anti-GD1b IgM antibody activity, and this finding correlated with vibratory loss on clinical exam. The authors also noted that GD1b IgM cross-reacted with anti GM1 IgM activity (27).
Anti-GM2 IgM autoantibodies. There are also cases reported of multifocal motor neuropathy associated with anti-GM2 antibodies, although these are rare and GM2 antibodies may be seen in other motor neuropathies and Guillan Barré syndrome (29; 121; 27).
Anti-MAG IgM neuropathy. The typical clinical picture of neuropathy associated with anti-MAG IgM antibodies is 1 of a distal, predominantly sensory or sensorimotor neuropathy, with demyelinating findings on nerve conduction studies, particularly in the most distal nerve segments. This clinical phenotype is quite distinct from that of patients with motor neuropathy or multifocal motor neuropathy. However, there have been patients described with anti-MAG antibodies and a predominantly motor neuropathy (04; 168). Presumably, patients with motor neuropathy and anti-MAG antibodies should be managed as would those with the typical MAG-neuropathy syndrome, although the number of cases is too small to adequately assess.
IgG and IgA monoclonal gammopathies. Patients with motor neuropathies and nonmalignant IgA monoclonal gammopathies have been reported (35; 187), including several with IgA lambda monoclonal gammopathies (19; 67; 117). Another patient had motor neuropathy with an IgGk monoclonal gammopathy associated with features of the Crow-Fukase syndrome including gynecomastia, hypertrichosis, leg edema, impotence, and a raised cerebrospinal fluid protein (16). It is not known whether the monoclonal gammopathy in these patients was coincidental or related to the neuropathy.
Amyloid. Patients with amyloid neuropathy typically have prominent sensory and autonomic symptoms at the onset, though motor involvement, including prominent fasciculations, frequently occurs. In 1 series, 2 of 18 patients had weakness as the first symptom (104).
A patient with a progressive lower motor neuron syndrome with leg weakness and fasciculations and a normal sensory examination and sensory nerve conduction studies, was diagnosed with amyloid neuropathy only after a motor nerve biopsy showing Congo-red staining in nerve and muscle (140). Although atypical, amyloid should be considered as a potential etiology of motor neuropathy.
Multifocal motor neuropathy has been reported as a rare adverse effect of infliximab, a monoclonal antibody drug that blocks tumor necrosis factor-alpha (12). A case report described multifocal motor neuropathy following treatment of ulcerative colitis with adalimumab, a similar anti-TNF-alpha agent (55). Adverse effects of immune checkpoint inhibitors have included demyelinating peripheral neuropathies, including acute inflammatory demyelinating polyneuropathy (105), polyradiculoneuropathy (126), and immune-related demyelinating peripheral neuropathy with conduction block (80).
A patient who was treated for serologically confirmed dengue infection developed asymmetric motor weakness of the hands more than feet without sensory symptoms, within 1 week after the dengue infection. Nerve conduction studies confirmed multifocal motor neuropathy with conduction blocks, and the patient responded well to intravenous immunoglobulin therapy (69). In a similar fashion, multifocal motor neuropathy has been described following Mycoplasma pneumoniae infection (116).
Limited epidemiological information is available for motor neuropathies. In a study of 169 patients with clinically defined motor neuron disease, conduction block was seen in 10 patients (6%) (95). Despite increasing awareness, multifocal motor neuropathy continues to be misdiagnosed as amyotrophic lateral sclerosis, accounting for 22% of misdiagnosed cases in Ireland from 1993 to 1997 (163). The incidence has been estimated at 1 to 2 per 100,000 population (118).
Among affected patients, epidemiology studies have reported that men are affected more often than women, at a ratio of 2.6:1. In 1 review of 294 patients reported in the literature, the mean age of onset was approximately 40 years of age, with most patients presenting between the ages of 20 and 50 years (118).
Patients with motor neuropathy are frequently suspected to have motor neuron disease because of the pure motor manifestations, the presence of muscle atrophy and fasciculations, and the frequent preservation of deep tendon reflexes. Clinical features that may distinguish motor neuropathy from motor neuron disease include the multifocal or asymmetric involvement and prolonged time course. Definite upper motor neuron signs such as a Babinski sign, clonus, or spasticity are not seen as a result of motor neuropathy. Electrophysiological studies may show persistent multifocal conduction blocks (133) or other signs of demyelination (33; 82). Highly elevated titers of anti-GM1 or anti-GD1a antibodies that are seen in motor neuropathy are not seen in motor neuron disease, although low level titers may be seen and are less specific (84; 157).
Motor neuropathy also appears to be distinct from chronic inflammatory demyelinating polyneuropathy. Both chronic inflammatory demyelinating polyneuropathy and multifocal motor neuropathy have evidence for demyelination and respond to immunosuppressive therapy (89), but they also have many distinguishing features. Unlike in chronic inflammatory demyelinating polyneuropathy, in motor neuropathy the nerve conduction velocities are usually normal between regions of block, sensory conduction studies are normal or mildly affected, cerebrospinal fluid protein is not commonly elevated, high titers of anti-GM1 or GD1a antibodies are frequently present, and the disease is rarely remitting. Nerve biopsy studies often show inflammation with chronic inflammatory demyelinating polyneuropathy, but not with motor neuropathy (42), and corticosteroids and plasma exchange are frequently helpful in chronic inflammatory demyelinating polyneuropathy, whereas they can exacerbate motor neuropathy (53; 45; 162).
Also to be considered in the differential diagnosis of multifocal motor neuropathy is Lewis-Sumner syndrome, also known as multifocal CIDP, or multifocal acquired demyelinating sensory and motor (MADSAM) neuropathy. This clinical entity has a multifocal and asymmetric pattern that resembles multifocal motor neuropathy; however, it typically has sensory involvement both clinically and on electrophysiologic testing. Additionally, this entity tends to behave clinically more like CIDP, with high CSF protein and similar response to immunotherapy. Anti GM1 IgM antibodies should not be present in Lewis Sumner syndrome (124; 102; 150; 178).
The diagnosis of multifocal motor neuropathy is based on a combination of clinical history and examination, serum studies, and electrodiagnostic findings. A number of consensus guidelines for criteria have been proposed in the diagnosis of multifocal motor neuropathy, including the European Federation of Neurological Societies/Peripheral Nerve Society (177) and the American Association of Neuromuscular and Electrodiagnostic Medicine (129). It should be noted that these are guidelines, and that there are patients who are clinically diagnosed with multifocal motor neuropathy and respond to IVIG therapy despite not meeting these criteria.
Core criteria (both must be present)
1. Slowly progressive or stepwise progressive, asymmetric limb weakness, or motor involvement having a motor nerve distribution in at least 2 nerves, for more than 1 month [usually more than 6 months].
2. No objective sensory abnormalities except for minor vibration sense abnormalities in the lower limbs.
Supportive clinical criteria
3. Predominant upper limb involvement. [At onset, predominant lower limb involvement accounts for nearly 10% of the cases.]
4. Decreased or absent tendon reflexes in the affected limb. [Slightly increased tendon reflexes, in particular in the affected arm, have been reported and do not exclude the diagnosis of multifocal motor neuropathy provided criterion 7 is met.]
5. Absence of cranial nerve involvement. [Twelfth nerve palsy has been reported.]
6. Cramps and fasciculations in the affected limb.
7. Upper motor neuron signs
8. Marked bulbar involvement
9. Sensory impairment more marked than minor vibration loss in the lower limbs
10. Diffuse symmetric weakness during the initial weeks
11. Laboratory: CSF protein >1 g/l
Electrophysiological criteria for conduction block
[Evidence for conduction block must be found at sites distinct from common entrapment or compression syndromes]
1. Definite motor Cba. Negative CMAP area reduction on proximal versus distal stimulation of at least 50%, whatever the nerve segment length (median, ulnar, and peroneal). Negative CMAP amplitude on stimulation of the distal part of the segment with motor CB must be greater than 20% of the lower limit of normal and greater than 1 mV (baseline negative peak), and an increase of proximal negative peak CMAP duration must be 630%.
2. Probable motor Cba. Negative CMAP area reduction of at least 30% over a long segment of an upper limb nerve with an increase of proximal negative peak CMAP duration 630%; or negative CMAP area reduction of at least 50% (same as definite) with an increase of proximal negative peak CMAP duration greater than 30%.
3. Normal sensory nerve conduction in upper limb segments with CB and normal sensory nerve action potential amplitudes (see exclusion criteria).
Criteria for definite multifocal motor neuropathy (129)
1. Weakness without objective sensory loss in the distribution of 2 or more named nerves. During the early stages of symptomatic weakness, the historical or physical finding of diffuse, symmetric weakness excludes multifocal motor neuropathy.
2. Definite conduction block (see AANEM 1999 criteria for conduction block below) is present in 2 or more nerves outside of common entrapment sites [median nerve at wrist; ulnar nerve at elbow or wrist; peroneal nerve at fibular head].
3. Normal sensory nerve conduction velocity across the same segments with demonstrated motor conduction block.
4. Normal results for sensory nerve conduction studies on all tested nerves, with a minimum of 3 nerves tested. The absence of each of the following upper motor neuron signs: spastic tone, clonus, extensor plantar response, and pseudobulbar palsy.
Criteria for probable multifocal motor neuropathy (129)
1. Weakness without objective sensory loss in the distribution of 2 or more named nerves. During the initial weeks of symptomatic weakness, the presence of diffuse, symmetric weakness excludes multifocal motor neuropathy.
2. The presence of either:
a. Probable conduction block in 2 or more motor nerve segments that are not common entrapment sites, or
b. Definite conduction block in 1 motor nerve segment and probable conduction block in a different motor nerve segment, neither of which segments are common entrapment sites.
3. Normal sensory nerve conduction velocity across the same segments with demonstrated motor conduction block, when this segment is technically feasible for study (that is, this is not required for segments proximal to axilla or popliteal fossa).
4. Normal results for sensory nerve conduction studies on all tested nerves, with a minimum of 3 nerves tested.
5. The absence of each of the following upper motor neuron signs: spastic tone, clonus, extensor plantar response, and pseudobulbar palsy.
Criteria for partial conduction block, as adapted from the AANEM (128)
Diagnostic studies that may be completed to rule out other etiologies of diseases presenting in a similar fashion to multifocal motor neuropathy include EMG/NCS (see criteria above); cerebrospinal fluid (protein > 1g/l is not consistent with multifocal motor neuropathy and suggests CIDP or Lewis Sumner syndrome); nerve biopsy; and serum studies, including SPEP/IFE, thyroid function testing, and CK (van Shaik et al 2006).
Blood/serum studies. Routine blood studies are normal in motor neuropathies. Estimates of the frequency of increased IgM anti-GM1 antibody titers in patients with motor neuropathy and multifocal conduction block range from 18% to 84% (84). A modified technique using an ELISA plate, which covalently links GM1 may increase the sensitivity of the assay (135), though the increase in sensitivity for multifocal motor neuropathy is minimal (24). In most cases, the IgM anti-GM1 antibodies are polyclonal, but in some they are monoclonal, although total serum IgM concentration is often increased. In 1 series, 4 of 14 patients with highly elevated IgM anti-GM1 antibody titers were monoclonal, all of which were nonmalignant; the rest were polyclonal (84).
Electrophysiological studies in patients with motor neuropathy and increased titers of anti-GM1 antibodies show varying degrees of abnormalities. In a review of patients with highly elevated anti-GM1 antibody titers, all with evidence of axonal degeneration and denervation, 5 of 14 patients had a single conduction block, 4 of 14 patients had multiple conduction blocks, and 1 patient had diffusely slow motor nerve conductions. In 4 of the 14 patients, conductions were normal, except for reduced compound motor action potential amplitudes, and the diagnosis would have been missed if not for the elevated anti-GM1 antibody titers. In 1 study, the presence of anti GM1 antibodies correlated with more severe disability and axonal loss in a cohort of patients with multifocal motor neuropathy (27); however, the correlation of GM1 antibodies to disease severity has been variable.
The finding of low titers of anti GM1 antibodies is not, however, specific for multifocal motor neuropathy. A number of cases have been reported of GM1 IgM antibodies in patients with motor neuron disease and other immune-mediated neuropathies, however, typically with lower titers (147; 93; 98; 176; 113).
Electrodiagnostic studies. In both motor neuropathy and motor neuron disease, electrophysiological studies may show reduced compound motor action potentials with EMG findings of fibrillations, fasciculations, and long duration neurogenic motor unit potentials consistent with axonal degeneration and denervation. In motor neuropathy, however, demyelinating features may be present to help distinguish it from motor neuron disease. These include significant slowing of nerve conduction velocities, prolongation of distal motor latencies or F wave latencies into a demyelinating range, temporal dispersion, or the presence of focal motor conduction block. Some patients with motor neuropathy exhibit minimal or no conduction abnormalities and may be difficult to diagnose.
In multifocal motor neuropathy, diagnostic criteria (as recommended by the EFNS/PNS and AANEM) depend on the finding of multifocal conduction block in motor nerves outside the typical compression sites. The definition of conduction block varies depending on which criteria are used, but the most conservative generally require 40% to 50% loss of CMAP amplitude between proximal and distal sites without significant temporal dispersion. Some patients with clinical multifocal motor neuropathy and positive GM1 antibodies may not have conduction block on nerve conduction studies but still respond to IVIG therapy. It is possible that these patients have very proximal or distal conduction block that cannot be detected with routine nerve conduction studies (130; 81; 119; 44; 153).
Conduction block. Persistent motor conduction block is 1 feature that unequivocally identifies a patient as having a neuropathy, though its demonstration is full of technical difficulties and challenges. Paranodal and internodal demyelination increases the transverse capacitance and reduces the resistance at the internode. This increases outward leakage current, increasing the time the internal longitudinal current must flow in order to generate an impulse at the next node of Ranvier. If the transverse current leakage is excessive, not enough current may be available to depolarize the next node of Ranvier and the impulse blocks (85; 23).
A drop in the compound motor action potential from proximal to distal stimulation is insufficient, by itself, to conclude the presence of conduction block. Apparent conduction block can result from submaximal stimulation at sites where the nerve trunk is deeply located, such as Erb point (40). Interphase shift and cancellation of the negative phase of a motor unit potential with the positive phase of another motor unit potential can result in significant drops in amplitude, particularly with chronic partial denervation. Papers describing patients with multifocal motor neuropathy and conduction block have variable definitions of conduction block ranging from 20% to 50% amplitude loss, or also requiring 50% area loss (136; 07; 172; 88; 148; 95; 02; 08; 33; 20). Amplitude loss of up to 41% and area loss of up to 29% may be seen in normal subjects (125). Computer modeling indicates that temporal dispersion can result in a drop in amplitude of greater than 50%, but a drop in area of greater than 50% is due to conduction block (142). In addition, short segment stimulation may be helpful in demonstrating conduction block, whereas a gradual change in amplitude is more likely to be caused by temporal dispersion (41).
Conduction block may be verified by comparing the partially blocked compound motor action potential with the surface recorded potential obtained during maximal volitional effort. This is accomplished by triggering on the largest peak of the recruitment pattern during maximal effort (96). If the volitional summated potential is larger than the proximal compound motor action potential, then true conduction block has not occurred, and the reduced proximal compound motor action potential may be due to submaximal stimulation.
Other demyelinating findings. Patients with motor neuropathies, diagnosed by the presence of conduction block or elevated titers of anti-GM1 antibodies, frequently exhibit other conduction abnormalities. In 9 patients with motor conduction block, the following additional features of demyelination were seen in the same or other nerves; temporal dispersion was seen in 5 patients, slowed motor conduction velocity in 7 patients, and prolonged distal motor latencies in 4 of the patients. All had prolonged F waves in at least 1 nerve (33). In another study of 16 patients with multifocal motor neuropathy, 9 patients with elevated anti-GM1 antibodies and 5 patients with conduction blocks, 15 patients had other features of demyelination; 5 patients had prolonged distal latencies, 7 patients had abnormal temporal dispersion, 8 patients had prolonged F-waves not explained by distal slowing, and 13 patients had conduction velocity slowing. Eight patients had at least 1 nerve with pure axonal features but had other nerves meeting demyelinating criteria. The 1 patient without meeting demyelinating criteria, had a prolonged F wave latency and a 31% drop in the median compound motor action potential area across the forearm. In patients treated with intravenous gamma globulin, improvement was observed in two thirds of the patients with conduction blocks, and in two thirds without conduction block (82).
Sensory nerve conduction studies. When sensory conduction has been assessed across the site of motor conduction block, either by assessing the ascending compound nerve action potential or using the near nerve technique to assess the ascending sensory nerve action potential, there has been normal sensory nerve conduction (133; 88). A preserved distal sensory response as is typically done does not ensure that the sensory conduction across a more proximal site of motor conduction block is conserved and that pure multifocal motor conduction block is present (87). However, patients with multifocal motor conduction block and distal preserved sensory responses do resemble clinically the patients in whom pure multifocal motor conduction block has been established (132).
Cerebrospinal fluid. CSF in patients with multifocal motor neuropathy typically shows a normal or mildly elevated protein; levels greater than 1 g/l make other etiologies, such as CIDP or MADSAM, more likely diagnoses. Oligoclonal bands are not typically present in multifocal motor neuropathy, and the IgG index is normal (20).
Nerve biopsy. There is usually no role for nerve biopsy in the typical evaluation and diagnosis of multifocal motor neuropath. In the appropriate clinical scenario, in which vasculitis or amyloid neuropathy are being considered, there may be a role for nerve biopsy to evaluate for these diagnoses; however, they typically have a different clinical presentation and include sensory symptoms and findings.
Magnetic resonance imaging. Some studies have reported T2 hyperintensity or contrast enhancement in MRI of the brachial plexus in patients with multifocal motor neuropathy (175; 173; 165). Although this finding has not been extensively studied and it is a nonspecific finding also reported in CIDP, it should be known that these radiographic findings are compatible with the clinical diagnosis of multifocal motor neuropathy, and this modality warrants further study as a diagnostic tool. In a series, asymmetric radiological abnormalities by brachial plexus MRI predominated in multifocal motor neuropathy or Lewis-Sumner syndrome, whereas symmetrical abnormalities were associated with CIDP (73). MRI has been demonstrated to show thickening along with altered diffusion in the median and ulnar nerves in multifocal motor neuropathy patients (62).
Ultrasonography. Multifocal nerve enlargement detected by ultrasonography has been described in patients with multifocal motor neuropathy (13). Sonographic enlargement of the brachial plexus and proximal median nerve segments is a specific finding for a chronic inflammatory neuropathy, in the absence of clinical features suggesting a hereditary demyelinating neuropathy (60). Imaging with ultrasound of the median and ulnar nerves in the forearm is potentially a powerful method to differentiate multifocal motor neuropathy from amyotrophic lateral sclerosis (74). Furthermore, there is evidence that ultrasonography may serve as a useful complementary diagnostic tool, as it may identify treatment-responsive patients with clinical phenotypes of an inflammatory demyelinating neuropathy who lack electrodiagnostic features of demyelination (59). However, there has been poor correlation between ultrasound findings and functional disability in multifocal motor neuropathy (83).
Intravenous immunoglobulin. The treatment of choice for motor neuropathy is intravenous immune globulin, which has been shown to be effective in placebo-controlled trials (08; 169; 52; 100). A favorable response has been reported in 67% to 100% of patients (32; 09; 153; 28), and a reduction of the degree of conduction block or an increase in the compound motor action potential may be seen (33; 36, 115; 52). Longer disease duration prior to the initiation of IVIG therapy has been associated with poorer outcomes (28).
Evidence-based guidelines published by the European Federation of Neurological Societies/Peripheral Nerve Society as well as the American Association of Neuromuscular and Electrodiagnostic Medicine (AANEM) recommend the use of IVIG as first line therapy for multifocal motor neuropathy, with recommended initial doses of 2 g/kg over 2 to 5 days followed by maintenance doses of 1 g/kg every 2 to 4 weeks or 2 g/kg every 4 to 8 weeks (177; 46). A 2018 open phase 3 multicenter trial demonstrated the safety and sustained efficacy over 52 weeks of an IVIG regimen of 1 g/kg every 3 weeks for treating multifocal motor neuropathy (92). An alternative dosing guideline has been proposed to more quickly optimize dosing and minimize costs (106). An initial IVIG dose of 2 g/kg is administered, with a second 2 g/kg dose administered 6 weeks later, if there has been an incomplete recovery. Subsequently, the patient is monitored until deterioration is noted, and this duration establishes the “dose interval.” Two more rounds of IVIG at 2 g/kg are given at this dose interval, with subsequent reduction in the dose by 20% per treatment until a relapse occurs, in order to determine the minimally effective dose.
The therapeutic effect of intravenous immune globulin is transient, and prolonged treatment is usually required. However, remissions either following the use of intravenous immune globulin (09), or occurring spontaneously (30), have also been reported. Benefit has been reported for periods of up to 4 years (09), and often long-term treatment is required (101). Occasionally, patients can have a diminished response after prolonged treatment (48; 09; 167), at which point other immunomodulating therapies, discussed below, may need to be considered.
Of patients with motor neuropathy and anti-GM1 antibodies, those with conduction block may have a better response to intravenous immune globulin (08), and of those with multifocal motor neuropathy with conduction block, the presence of anti-GM1 antibodies may be associated with a better therapeutic response (20; 09; 112). However, other studies have suggested that the presence or absence of GM1 antibodies or conduction block does not predict response to therapy (153), and many studies report that patients without anti-GM1 antibodies or conduction block also respond to intravenous immune globulin (37; 82; 130; 49). GM1 antibody titers appear not to correlate with clinical improvement and should not be followed as a disease marker (36).
The mode of action of intravenous immune globulin in motor neuropathy is unknown, and it may exert its effects by several different mechanisms. Titers of anti-GM1 antibodies remain unchanged after intravenous immune globulin treatment, but the intravenous immune globulin may exert anti-idiotypic activity (108; 189), or it may block Fc receptor-mediated recruitment of macrophages, complement activation, (191), or action of various cytokines, as has been suggested for other autoimmune diseases (90; 01; 11). One in vitro study of GM1 IgM activation showed IVIG to block complement activation, suggesting another mechanism of therapeutic action (179).
Patients with autoimmune neuropathies such as multifocal motor neuropathy have been justifiably concerned about whether their disease status increases the risk of infection with COVID-19. There is currently no evidence that multifocal motor neuropathy itself increases susceptibility to COVID-19. Therefore, if a patient is clinically stable, there are no compelling or data-driven reasons to change anything in these patients and disturb clinical stability. For patients on monthly IVIg, there may even be a theoretical advantage that IVIg offers additional protection due to natural autoantibodies; if IVIg is not infused as home infusion, switching to self-administered subcutaneous IgG might be an option to diminish exposure (43).
Subcutaneous immunoglobulin. There has been growing evidence that patients maintained on IVIG for multifocal motor neuropathy may be successfully maintained on subcutaneous formulations. In 2 case series of IVIG-responsive patients, treatment with subcutaneous IVIG demonstrated overall stable clinical course and well tolerated treatment (47; 63). There have since been limited studies of subcutaneous administration of immunoglobulin suggesting preservation or improvement of muscle strength, ability, and quality of life, with fewer systemic side effects and superior cost-effectiveness compared to IVIG therapy (110). A meta-analysis of 8 studies totaling 50 patients with multifocal motor neuropathy found no differences in muscle strength outcomes in multifocal motor neuropathy when treated with subcutaneous immunoglobulin compared to IVIG, with a 28% reduction in relative risk of moderate and/or systemic adverse effects (141). A prospective open-label study comparing efficacy of IVIG to facilitated subcutaneous immunoglobulin in 18 patients similarly found no significant differences in muscle strength and disability (68). Other small studies have demonstrated that subcutaneous immunoglobulin is feasible and effective for the long-term maintenance treatment of patients with multifocal motor neuropathy (03; 58). Further larger randomized trials must be conducted to better evaluate the safety and efficacy of subcutaneous IVIG.
Steroids and plasma exchange. Oral and intravenous corticosteroids, plasmapheresis, and immunoadsorption have generally been ineffective in the treatment of motor neuropathies (79; 78; 170). In addition, corticosteroids and plasmapheresis have been associated with disease exacerbations (53; 45; 162; 25; 54). As clinical worsening with corticosteroid or plasma exchange therapy has been reported specifically in multifocal motor neuropathy, the distinction of this clinical syndrome from CIDP or Lewis-Sumner syndrome is an important one.
Rituximab. Studies of rituximab, a monoclonal antibody against the B cell marker CD20, have been promising. In a 2003 study of rituximab for a number of neuropathies, 14 patients with multifocal motor neuropathy receiving rituximab had improved strength compared to 8 patients not receiving the medication (137). A 2009 report of 3 patients with decreasing response to IVIG showed sustained clinical improvement following rituximab therapy (154). A 36-year-old man who had suffered a slow decline despite chronic IVIG treatment had significant improvement in strength after rituximab, with a stable course without further treatment over the following 8 years (61). In addition to these studies, there are other reports of single cases of multifocal motor neuropathy responding to rituximab treatment. These studies suggest a role for rituximab in patients in whom there is decreasing response to IVIG. However, in a small open label trial of 2 doses of rituximab given to 6 patients in conjunction with IVIG therapy, there was no reduction in the amount of IVIG given or in scores of strength or disability. The study was limited by small size and concurrent IVIG treatment, and none of the patients had GM1 antibodies (31). Further, larger studies are needed to better assess the efficacy and safety of rituximab in patients with motor neuropathy and multifocal motor neuropathy, but initial reports are promising that it may be an effective treatment in patients who become unresponsive to IVIG.
Cyclophosphamide. Motor neuropathy has also been reported to respond to the chemotherapeutic agents chlorambucil (98), intravenous cyclophosphamide (136; 87; 53), and fludarabine (152), all of which lower autoantibody titers and serum IgM concentrations and reduce dependency on intravenous immune globulin. Of these, cyclophosphamide is the most established and often described.
A number of other case reports and series have described improvement in multifocal motor neuropathy refractory to IVIG therapy with oral or intravenous cyclophosphamide. There are cases of oral cyclophosphamide being added to IVIG, allowing a decrease or cessation of IVIG treatment (115). Others have improvement in some patients on chronic IVIG with the addition of intravenous cyclophosphamide (09). One patient with refractory multifocal motor neuropathy improved after treatment with high dose intravenous cyclophosphamide without stem cell rescue (22). In 1 review of 25 patients with multifocal motor neuropathy, 8 required adjunctive immunotherapy, and in some of these cases, this included cyclophosphamide (101). Although there are no randomized controlled trials, in those patients for whom additional immunotherapy is required when there is no longer a response to IVIG, cyclophosphamide should be considered.
Interferon beta 1a. Two trials of have examined the use of beta interferon 1a in multifocal motor neuropathy. One study demonstrated initial improvement in the first 3 months of treatment and then clinical stabilization (111). In another study, 3 of 9 patients treated had improvement (174). These studies are small, and the use of beta interferon 1a in multifocal motor neuropathy remains a question.
Mycophenolate. Although a small initial study demonstrated IVIG dose reductions in patients with multifocal motor neuropathy treated with mycophenolate (15), a more recent randomized controlled trial of mycophenolate versus placebo added to IVIG showed no benefit in terms of IVIG reduction, strength, functional scores, or GM1 IgM titers (139). It is possible that a longer or larger study is required to better assess the role of mycophenolate in the treatment of motor neuropathies.
Azathioprine. There have been single reports of individual patients with multifocal motor neuropathy responding to azathioprine treatment after failure of steroids or IVIG (64; 20). Other reports have described no benefit (88), and no larger series or systemic trial of azathioprine for multifocal motor neuropathy has been published.
There is limited information about the course and treatment of multifocal motor neuropathy during pregnancy. There have been limited reports of worsening of multifocal motor neuropathy symptoms during pregnancy, requiring treatment with IVIG (34). Given the low prevalence of this disorder among young women, there is limited epidemiologic and treatment management information.
All contributors' financial relationships have been reviewed and mitigated to ensure that this and every other article is free from commercial bias.
Robert W Pratt MD
Dr. Pratt of the University of Colorado has no relevant financial relationships to disclose.See Profile
Louis H Weimer MD
Dr. Weimer of Columbia University has received consulting fees from Roche.See Profile
Nearly 3,000 illustrations, including video clips of neurologic disorders.
Every article is reviewed by our esteemed Editorial Board for accuracy and currency.
Full spectrum of neurology in 1,200 comprehensive articles.
Listen to MedLink on the go with Audio versions of each article.
Jan. 23, 2023
Jan. 12, 2023
Jan. 05, 2023
Jan. 04, 2023
Jan. 04, 2023
Jan. 03, 2023
Jan. 03, 2023
Jan. 03, 2023