by Julius Birnbaum, MD
Fellow, Division of Rheumatology
Johns Hopkins University School of Medicine
Release Date: September 26, 2006
Expiration Date: September 26, 2008
Dr. Birnbaum has no significant financial interest or relationships to disclose.
Today’s discussion will focus on three cases with neurologic findings that were encountered during the course of a first-year rheumatology fellowship. The trajectory of the cases will graduate from syndromes with well-described mechanisms and incontrovertible findings to syndromes where the etiology of the neurological disease is less clear. What I will try to illustrate is that the border between rheumatic disease and primary demyelinating neurological disease is an area that is pockmarked with nebulous disease mechanisms, with corresponding tentativeness, if not outright confusion, with regard to treatment.
The first patient is a 72-year-old African American man with a history of non-insulin dependent diabetes, hypercholesterolemia, and psoriasis. He had been on Lipitor for 5 years. In October 2005, he was referred for left-side weakness that had been worsening since March, 2005. In March and April, he had experienced transient numbness in the left lower extremity that resolved. However, weakness progressed, with symptoms of left hemiparesiss affecting the left leg more than the left arm, and with sparing of the face. Some of his friends had said that they thought he had suffered a stroke because he had tonic flexor posturing of the left arm. He was referred to Johns Hopkins because his primary care doctor drew labs and found a creatinine kinase (CK) of 1025. He had a normal MRI/MRA of the brain. The question to us was: “Can you get a lateralizing myopathy?” He denied any proximal motor symptoms on the right arm or the right leg. There were no bulbar or sphincteric symptoms.
Results of his neurological exam were as follows :
- Anisocoria: The right pupil was 4 millimeters, and was reactive to 2 milimeters. The left pupil was smaller, 3 millimeters reactive to 2. The suggestion that this anisocoria was non-physiologic was borne out by the finding that he had mild left eye ptosis.
- Facial sensation and strength was symmetric.
- He had what is referred to as a “pyramidal” pattern of motor weakness: There is disproportionate weakness in the extensors of the upper extremity, relative to the flexors. His triceps and wrist extensors were weakest in the upper extremity. The opposite is the case in the lower extremity. Pyramidal weakness affects flexors more than extensors. Therefore, the hip flexors and the hamstrings were the weakest, compared with the anti-gravity muscles. This suggests a weakness affecting the corticospinal tract, or so-called “ long-tract” findings.
- He had hyperreflexia more on the left than the right, which upholds the suggestion that he had a pathological process involving the corticospinal tracts.
- He had bilateral Babinski’s, which also supported the suggestion that this was a bilateral corticospinal process.
- He had a left hemisensory syndrome.
- His gait was also mildly spastic.
In summary, he had a left Horner’s syndrome, which spared the face and the bulbar musculature. He had bilateral, left-more-than-right, long-tract signs. In his motor exam, he had increased tone, and he had a pyramidal pattern of left-lateralizing weakness. Again, the extensors are affected more than the flexors in the upper extremity, with the opposite in the lower extremity. Additionally, reflexes are suggestive of a corticospinal tract dysfunction, lateralizing to the left side; the gait also lateralized to the left side.
The question is: “How can we localize this Horner’s syndrome, which spares the face and the bulbar muscles and which presents with left-lateralized, pyramidal distribution of weakness?”
Understanding the anatomy of the sympathetic outflow tract is crucial.
First-order sympathetic nuclei reside in the hypothalamus, ramify through the brain stem, go through the interomediolateral tracts of the spinal cord, and then finally synapse on T1. Lesions anywhere along this pathway can lead to Horner’s syndrome. In what is called first-order Horner’s syndrome , you get lesions from the area of the hypothalamus nuclei to the T1. Given that Horner’s syndrome is occurring in the context of bilateral corticospinal tract signs, the lesions of Horner’s syndrome must occur where the corticospinal tracts and the sympathetic outflow tracts are most closely juxtaposed in space. This occurs in the brain stem or the cervical cord.
So given that we know there is a Horner’s syndrome and bilateral pyramidal signs, we are suspicious for a lesion in either the brain stem or the cervical cord. And given that facial weakness is not observed, that pharyngeal and bulbar function is spared, this suggests that we have a lesion below the neck, affecting the cervical cord.
We are suspicious of a lesion on the so-called ciliospinal tract, or at least affecting that part of the cervical cord.
The other important finding that helps localize a lesion is the fact that our patient’s upper extremity reflexes are hyperreflexic. If he had a lesion that was affecting the biceps or the triceps, we would expect to see a dropped bicep reflex, a dropped triceps reflex. The fact that these reflexes are enhanced suggests that the lesion is actually occurring above C5.
The summary of this neurological exam is that we feel suspicious that there is a cervical lesion occurring below the medulla and above C5. When I saw this patient, I was concerned that he had a high cervical cord compression. An MRI was ordered. The MRI showed severe cord compression.
On the sagittal view, it looked like the diameter of the spinal cord was narrowed down to 2 or 3 millimeters. It also looked as if there was impingement from the posterior parts of vertebral bodies, suggesting that the posterior longitudinal ligament was contributing to cord compression.
In the axial view, the spinal cord was squeezed to the size of spaghetti. He also had a 3D reformatted CAT scan of the C-spine, which additionally showed ossification of the posterior longitudinal ligament contributing to cord compression.
An emergency referral to neurosurgery prompted an immediate multi-level laminectomy. His post-operative, of course, was terrific. According to the primary care doctor, he regained nearly full use of his arms and leg, and was walking without difficulty. With regard to his history of elevated CK, he has remained off cholesterol-lowering medicines, and his primary care doctor is not checking CK markers as long as he remains asymptomatic.
The take-home point from this first case is that the close juxtaposition of motor, sensory, and autonomic tracts in the spinal cord can give rise to a variety of syndromes, depending on the longitudinal and anteroposterior extent of the lesion. Root signatures of either hyperactive or dropped reflexes are a valuable guide in terms of localizing the lesion. (These general principles of a neurological exam will be useful in the next case.)
The next patient is a woman whom I met during the first month of inpatient consultations: tetraparesis in a woman with lupus. She was a 32-year-old African American woman who was diagnosed with lupus in 1997, when she presented with polyarthralgia, positive ANA, positive anti-double-stranded DNA, Coombs-negative anemia, and thrombocytopenia. From 1997 to 2004, she had fairly- well-controlled lupus, maintained on Plaquenil 400 milligrams per day, with varying doses of prednisone given at times of maximum polyarthritis. Her lupus was relatively quiescent until December 2004, when she presented to an outside hospital with complaints of intermittent spasms of the left lower extremity that progressed to monoparesis. An MRI of the spine revealed a Gadollinium-enhancing lesion from C3 to C5, as well as a non-enhancing lesion demonstrating T2 hyperintensity in the ventral part of the upper thoracic cord. A lumbar puncture revealed a neutrophilic pleocytosis, 97% neutrophils, 41 WBCs, with normal protein and glucose. Oligoclonal bands and IgG index were negative. The patient received her first course of pulse steroids, with complete return to her premorbid state.
In March 2005, she was hospitalized for workup of abdominal pain. An EGD showed an erosive gastritis. While hospitalized, she started complaining of numbness in both arms and both legs. Within days, she developed a quadriparesis. She developed right-more-than-left, upper extremity weakness with significant paraparesis, or weakness of the lower extremities. On neurological exams, her leg musculature was 2/5 MRC scale. More important, from a functional perspective, she lost complete ambulatory capacity, and required Foley catheterization. An MRI at this time revealed a cervical longitudinal myelitis. She was given her second course of pulse steroids, started on IV Cytoxan®, but she returned to only about 60% of her premorbid state, and now required a walker.
Then, in May 2005, after her second course of IV Cytoxan®, she developed recurrent paraparesis, with loss of independent ambulatory capacity. She was treated with a third course of IV steroids. Over weeks, she recovered to about 40% of her premorbid baseline. At this point, the Cytoxan® was discontinued because it was not known if this was transverse myelitis secondary to lupus, as her background lupus activity was otherwise quiescent.
In July 2005, she developed three new symptoms: subacute, mildly painful decreased acuity in the left eye; lack of coordination of the right hand; and worsening paraparesis. There were no other symptoms of active lupus disease. Her visual symptoms completely resolved in 1 week. In terms of her serology, she was only ANA-positive 1 to 320. The rest of the autoantibodies were negative. She was given a fourth course of pulse steroids, and was transferred to Johns Hopkins for further evaluation.
Her motor exam was as follows:
- Her upper extremities were normal.
- In the lower extremities (as in our first case), she had a “pyramidal” pattern of weakness affecting the flexors more than the extensors in the lower extremities, right more than left. But in contrast to the first patient, she did have multiple root signatures. She had completely absent or dropped biceps, triceps, brachioradialis, and finger flexor reflexes. This suggests that the myelitis was affecting the ventral, motor nuclei, in the gray matter, from C5 to T1.
- The fact that she had a myelopathy with long-tract signs is borne out by the fact that she had brisk reflexes with pathological spread, and bilateral Babinski’s.
- She had mild dystaxia, on right finger—to –nose testing.
- Her gait was broad-based and spastic. At the time we saw her, she was only able to ambulate one or two steps with the assistance of two people.
Her neurological exam suggests that we have bilateral pyramidal symptoms with increased lower extremity reflexes, suggestive of a myelopathy. However, the absence of reflexes in the upper extremities suggests there is a longitudinal myelitis affecting C5 through T1. In the context of the extensive longitudinal myelopathy, the episode of visual loss, although resolved, is still concerning for an optic neuritis.
In the T2 sagittal view of her MRI, there is patchy increased signal spanning the entire extent of the cervical cord down to the upper thoracic vertebral bodies. Radiographic evidence of inflammation in the spinal cord, which spans at least three vertebral segments, is referred to as “ longitudinal myelitis” .
If you look at the American College of Rheumatology classification criteria for neuropsychiatric lupus, the syndrome of myelopathy is nakedly enumerated, without an accompanying algorithm for distinguishing between a spinal cord disease caused by lupus and a primarily demyelinating syndrome such as multiple sclerosis is not provided. This distinction is crucial because interferons, which are used as a treatment for MS, can worsen lupus.
Key questions arise after we see that we have not only a pan-cervical myelopathy, as suggested by neurological exams, but a longitudinal myelopathy are:
- Is this transverse myelitis secondary to lupus?
- Or, is it a transverse myelitis secondary to MS or perhaps another so-called idiopathic primary demyelinating disease, such as neuromyelitis optica?
- What are the important pathological mechanisms that are driving this recurrent myelopathy?
Consideration of these questions is crucial, because such diagnostic and mechanistic uncertainty led to a confused treatment regimen, where Cytoxan and prednisone were haphazardly started and stopped, with significant progression of neurological deficits.
Two features of this case help narrow the differential diagnosis. One is the history of transient visual loss — important because, even if it had resolved when we met her, we need to focus on the differential diagnoses of opticospinal syndromes. This would include Devic’s syndrome, or so-called neuromyelitis optica (NMO), which is either idiopathic or secondary to connective tissue diseases; multiple sclerosis; or antiphospholipid antibody syndrome. Her antiphospholipid antibodies were negative.
Longitudinal myelitis should lead to consideration of Devic’s syndrome even before symptoms of optic neuritis. Devic’s syndrome, or neuromyelitis optica (NMO) , is a demyelinating opticospinal syndrome that is characterized by concomitant or sequential evolution of optic neuritis and transverse myelitis. Initial diagnostic criteria emphasized temporal clustering of attacks of optic neuritis with transverse or longitudinal myelitis, emphasized bilaterality of optic neuritis, and emphasized that NMO was monophasic Recent criteria place no restriction on the timeframe separating optic neuritis and transverse myelitis. It is recognized that optic neuritis can be unilateral, and a polyphasic course is now known to occur in up to 70% of cases.
If both MS and NMO can be characterized by these opticospinal symptoms of transverse myelitis and optic neuritis, what are clinical or imaging features that might validate NMO as a separate nosologic entity? The myelopathy of NMO is frequently more severe. It leads to very rapid functional decline with greater motor and sensory symptoms. A key finding is that, on MRI, you see a longitudinal myelitis, usually spanning more than 3 vertebral segments, with increased signal filling the entire anteroposterior width of the spinal cord. In contrast, the myelopathy of MS is less fulminant. It usually presents more insidiously. On MRI, either the inflammatory process is frequently focal, affecting less than 3 vertebral segments, and is usually restricted to the dorsolateral part of the cord.
There are also other distinctions. In MS, you rarely see the pleocytosis that you do in NMO. Formal diagnostic criteria describing MS have incorporated the presence of oligoclonal bands and IgG index. In contrast, in NMO, significant CSF pleocytosis is seen in up to to 75% of patients. Additionally, oligoclonal bands are seen in no more than 30% of patients. The other important distinction is that in NMO, the MRI of the brain is normal in 80% of the cases. The remaining 20% feature nonspecific white matter changes that are not pathognomonic or suggestive of MS lesions.
These distinguishing features have been formalized into diagnostic criteria. In evaluation of possible NMO, there needs to be absolute objective evidence of both optic neuritis and myelitis, with no evidence of disease outside the optic nerves and the spinal cord. For the diagnosis of NMO, at least one of the following supportive criteria needs to be present:
- Negative brain MRI
- Spinal cord MRI that shows a longitudinal myelitis
- CSF pleocytosis
For the second patient, neurological exam suggested a pancervical myelopathy. The MRI documented a longitudinal involvement of the spinal cord, from the cervical to the mid-thoracic level This longitudinal extension is very highly concerning for Devic’s syndrome. Despite a normal ophthalmologic exam, we have a strong suspicion that our patient had an episode of optic neuritis, probably because of retrobulbar neuritis. When the optic nerve is affected behind the disc, you often do not necessarily see acute changes, such as papillitis, at the level of the disc.
Apart from a clinical and radiographic distinction, there is also histopathological evidence that NMO is a distinct syndrome. Autopsy studies have shown that, in contrast to MS, which features T-cell perivenular demyelinated plaques, there is a vasculocentric pattern of immunoglobulin and complement deposition along hyalinized microvessels in the spinal cord, suggesting that NMO might be a humorally-mediated process affecting the spinal cord vasculature. The scary histopathologic feature is that lesions showing active immunoglobulin and complementary deposition can be seen next to other lesions with showing coalescing, necrotic cord lesions of gray and white matter.
The important point here is to initiate treatment at a time when the disease is marked by an inflammatory phase, as this might be important in preventing cylindrical and irreversible necrosis of gray and white spinal cord tracts, irreversible visual loss, and loss of functional ambulation.
She received six months of Cytoxan. There was no improvement in her symptoms. An MRI of the C-spine showed a static lesion load, similarly without any improvement. A report noted “an inability to establish a definite diagnosis of multiple sclerosis.” The plan was to re-evaluate the patient in the Rheumatology Clinic and start a trial anticoagulation, with frequent clinical and radiographic evaluation.
Because we were suspicious for NMO, the patient was treated with five courses of plasmapheresis, and started on Cellcept. There is no consensus regarding maintenance therapy for NMO. Nevertheless, she he had a dramatic response. Within 1 week, she had significantly improved paraparesis and was able to ambulate up to 30 steps with minimal assistance.
What happened afterwards was unfortunate. She was discharged on Cellcept, but this was discontinued because of GI symptoms. She was intermittently compliant, and received only two further doses of IV Cytoxan. In December 2005, she was readmitted for worsening paraparesis. However, she concomitantly had bilateral decreased visual acuity, to the point of being almost blind. Ophthalmologic evaluation finally revealed optic neuritis. With the objective physical findings of optic neuritis, the patient finally satisfied formal criteria for NMO. There was concern for rostral migration of her myelopathy, which could potentially lead to diaphragmatic weakness.
This is a case where you ask, “How do you tell there is active disease?” It is definitely a challenge. The NMO IgG autoantibody serves as a marker, but it also might serve as a valuable prognostic assay when you have lupus patients presenting with transverse myelitis. We could anticipate that a myelopathy might progress to clinically devastating involvement of the optic nerves. In fact, she is now left with severe decreased visual acuity.
The take-home point from this case is that clinical and radiographic detection of the discriminating features of a longitudinal myelitis, when coupled with presence of positive IgG NMO autoantibody, should allow for treatment for the aggressive nature of NMO syndrome, even before the emergence of optic neuritis.
The third and last patient is a 57-year-old woman with a 15-year history of multifocal sharp and ancinating pain. Her history includes multifocal, poorly localizable, sensory symptoms described as “stabbing sensations,” occurring in the right hand, but sometimes involving the right leg. When this first started, the pains lasted only seconds. However, over the course of 15 years, the sensations now lasted up to days. She was seen by multiple doctors. An MRI of the C-spine showed degenerative disc disease, which prompted a neurosurgeon to recommend a multilevel laminoplasty. She was seen by an outside neurologist, and found her way to the Johns Hopkins Multiple Sclerosis Center.
At the Multiple Sclerosis Center revealed evidence of a cervical myelopathy with a pyramidal pattern of mild paraparesis, , lower extremity reflex preponderance, and right-sided Babinski. A review of the MRI showed a high signal intramedullary cord lesion which extended above the area of spondylytic ridging, suggesting that the entire myelopathy might not be driven by structural disc disease. At first, it t was felt that the intramedullary signal might represent an idiopathic and incomplete transverse myelitis. However, given her chronicity of multifocal sensory symptoms, with only very mild paraparesis, it was decided that treatment did not need to be started at that time.
In July 2004 she had a serological evaluation that illustrated a significantly elevated RVVT o 49.3 seconds, which did not correct on mixing; presence of lupus anticoagulants; and an elevated anticardiolipin IgM. She reported no symptoms of Raynaud’s. However,, in August, she had an acute onset of bluish, painful discoloration in her right thumb, which lasted 10 days, and then resolved. She did not go to a physician, as she had become disenchanted with the medical profession, because they could not come up with a diagnosis. She did have a follow-up echocardiogram, which revealed no vegetation. She was lost to follow-up, but then presented almost 8 months later to the MS Center, in March 2005. She came in in a wheelchair. The exam suggested that there was progression of cervical myelopathy. An MRI now showed radiographic progression of the cervical cord lesion, to the extent that it was thought now that she might have multiple sclerosis. She was started on Cytoxan.
But, what about these antiphospholipid antibodies?
Her MRI revealed very patchy, discrete lesions that do not extend below the cervical spine. They do not occupy the entire ventroposterior span of the cord; it showed increased signal occupying only a minimal part of the cord. These lesions are characteristic of MS lesions. In the brain, the corpus callosum is one of the areas disproportionately affected in MS. The rest of the brain has nonspecific white matter lesions, not at all characteristic of MS. Even though the neurologists had committed to treatment for MS, they were concerned about the disease course, which is atypical even for a subtype of MS known as primary progressive MS.
Referral to Rheumatology
Concerned about the presence of antiphospholipid antibodies, they referred her to the Rheumatology Clinic, to see whether there was any underlying inflammatory syndrome. Her review of symptoms was negative. Her general physical exam was unremarkable. The neurological examination confirmed evidence of a cervical myelopathy, with right dorsiflexion weakness; long-tract signs were confirmed by reflex preponderance in the lower extremities, with the presence of bilateral Babinskis. Her antiphospholipid antibodies were significantly positive — RVVT of 47, beta-2 glycoprotein IgM of 82.
What are some of the neurological manifestations of antiphospholipid antibody syndrome? What criteria allow distinction from MS?
Briefly, MS is believed to be a principally T-cell mediated syndrome. About 85% of the time, MS progresses through relapses, with resulting morbidity due to axonal loss and scarring. Diagnostic criteria have undergone several revisions. What has not changed is the orthodoxy that you need events which are disseminated in time and space. Initially, before MRI, this needed to be fulfilled by history and physical exam. Now, the ability of MRIs, CSF, and Visual Evoke Potential to serve as surrogate markers of inflammation has allowed these paraclinical tools to be incorporated into formal diagnostic criteria. This allows for earlier diagnosis when you have a single clinical isolated demyelinating episode.
As mentioned above, about 85% of patients progress through a relapsing-remitting phase. However, 10% to 15% of patients have an indolent, chronic progressive course. Usually it presents as a myelopathy without any superimposed flares of brainstem disease. This falls under the rubric of “primary progressive disease.” The diagnosis of a primary progressive MS clinically requires at least one year of indolent disease progression.
So does our patient have primary progressive MS?
Does she present with at least one year of disease progression? Yes.
But she does not have a positive brain MRI, defined as 9 T2 lesions, or 4 more T2 lesions with a positive Visual Evoke Potentials. She has a positive spinal cord MRI; however, she does not have any evidence of intrathecal antibody production. Even though she has several white matter lesions that are similar to MS, she does not satisfy the criteria for MS.
Can a myelopathy be occurring in the context of antiphospholipid antibody syndrome? The problem is that it is difficult to determine whether antiphospholipid antibodies have an etiological or an associative role. This is the gray area of the neurological manifestations of APL. However, it’s been shown that in certain syndromes that cannot be otherwise explained, you see an increased incidence of antiphospholipid antibodies.
The question is, when is MS really MS? When is atypical MS really due to antiphospholipid antibodies? There is no good answer. In patients with atypical presentations of MS, do we need to consider the possibility that it is a neurological manifestation of APL, and treat with anticoagulation? The studies on this are not robust as there is no uniformity of patient population, different assays for APL are being tested, and, in most cases, elevated antiphospholipid antibodies likely represent an epiphenomenon of heightened immune activation rather than serving an etiological role. I think that in atypical cases such as this one, the testing for APL is probably worthwhile. If the patient does not respond to conventional MS treatment, a trial of anticoagulation should be considered.
How can neurological manifestation of APL be distinguished?
Optic neuritis in APL is usually more acute. It is rarely bilateral. The CSF and antiphospholipid antibody syndrome lacks lymphocitic pleocytosis and lacks evidence of intrathecal antibody production as measured by oligoclonal bands and elevated IgG index. They also differ radiographically.
In summary, these cases illustrate how similar clinical syndromes, characterized by cervical myelopathy, can give rise to a rich array of diagnostic possibilities that fall at the intersection of neurologic and rheumatic disease. The first case showed the importance of discriminating between a primary rheumatologic and neurologic disorder. The second case illustrated the importance of using neurological examination to further distinguish between neurologic manifestations of rheumatic disease, i.e., NMO and primary demyelinating disorders such as MS. The third case illustrated that you have to have a high level of tolerance for diagnostic uncertainty at this intersection. In these areas — where primary neurologic disease and neurologic rheumatic disease gingerly intercept — the mechanisms are not at all clear. It remains an exciting area for mechanistic studies.