Clinical and MRI correlates in 27 patients with acquired pendular nystagmus.

Clinical and MRI investigations were carried out on 27 patients with acquired pendular nystagmus in an attempt to delineate possible sites of lesions responsible for pendular nystagmus and mechanisms underlying the frequent ocular disconjugacy of this nystagmus. The aetiologies were multiple sclerosis (n = 21), brainstem stroke (n = 3) and other neurological conditions. In at least 59% of the patients, pendular nystagmus appeared > 1 year after the first symptom of the disease. Patients MRIs were characterized by multiple areas of abnormal signal and were analysed statistically to identify areas where lesions overlapped significantly between patients. Statistically significant overlap occurred in areas containing the red nucleus, the central tegmental tract, the medial vestibular nucleus and the inferior olive. Patients with horizontal pendular nystagmus showed predominantly pontine lesions whereas patients with torsional pendular nystagmus showed predominantly medullary involvement. The nystagmus was conjugate in 15 patients and disconjugate in amplitude or direction in 12. Internuclear ophthalmoplegia or asymmetrical visual acuity occurred in similar proportions in both groups. Patients with conjugate pendular nystagmus had a higher incidence of symmetrical, "mirror image' lesions on MRI than patients with disconjugate nystagmus. The abundance of abnormal MRI signals in our sample suggests that large or multiple structural lesions may be required to elicit pendular nystagmus, predominantly in the pons but also in the midbrain and medulla. The involvement of structures projecting to the inferior olive supports the hypothesis that oscillatory properties of olivary neurons causes the rhythm of pendular nystagmus. The delay observed between the onset of the underlying disease and the pendular nystagmus supports a mechanism operating via neural deafferentation. Disconjugancies in pendular nystagmus cannot be explained on the basis of the associated internuclear ophthalmoplegias nor on the basis of asymmetrical visual acuity. The association between symmetrical MRI lesions and conjugate nystagmus suggests that asymmetrical damage to brainstem structures concerned with binocular alignment may underlie disconjugate pendular nystagmus.

Acquired pendular nystagmus: oculomotor and MRI findings.

The clinical, oculomotor and ophthalmological features of 27 patients with pendular nystagmus were studied in whom 22 also had MR imaging of the brainstem. The nystagmus was predominately horizontal in 4 patients, torsional in 5, vertical in 3 and mixed in trajectory in 8. Fifteen patients had conjugate nystagmus. Twelve patients had disconjugate nystagmus. Eight patients had INO. In 16 patients visual acuity was 6/12 or worse. Acuity and the presence of INO were unrelated to the conjugacy of the nystagmus. The MRI cuts at the medullary, pontine and midbrain levels were analysed statistically to determine the areas where there was significant (< 0.05%) overlap between areas of abnormal signal in different patients. Significant target areas for lesions causing the nystagmus were: in the pons the medial vestibular nucleus, central tegmental and paramedian tracts; in the medulla the inferior olivary nucleus, reticular formation, dorsal accessory olivary nucleus, central tegmental tracts and olivo-cerebellar fibres; in the midbrain the red nucleus and central tegmental tracts. Horizontal pendular nystagmus was preferentially associated with pontine lesions and torsional nystagmus with medullary lesions. Patients with conjugate nystagmus had a tendency to have bilateral mirror image MRI lesions (p = 0.028). The prevalence of lesions in our patients raises a possibility that more than one neuronal mechanism must be affected to produce pendular nystagmus. The inferior olive may be responsible for the rhythm of ocular oscillation. The disruption of pathways proximal to the oculomotor nuclei may determine the instability in terms of individual eye movement.

Magnetic resonance imaging in degenerative ataxic disorders.

MRI of the brain was performed in 53 patients with a variety of degenerative ataxias and related disorders and 96 control subjects. Atrophy of intracranial structures was not seen in patients with the pure type of hereditary spastic paraplegia, or in early cases of Friedreich's ataxia. In advanced Friedreich's ataxia there was atrophy of the vermis and medulla. The MRI features of early onset cerebellar ataxia with retained reflexes were variable, and suggest heterogeneity. In autosomal dominant cerebellar ataxias, most patients had cerebellar and brainstem atrophy, probably reflecting the pathological process of olivopontocerebellar atrophy; there was no clearly defined group with both clinical and imaging features of isolated cerebellar involvement. The MRI abnormalities in idiopathic late onset cerebellar ataxia were predominantly those of cerebellar and brainstem atrophy or pure cerebellar atrophy. The clinical and imaging features of brainstem abnormalities were discordant in several patients. Pure cerebellar atrophy was associated with slower progression of disability. Cerebral atrophy was common in the late onset ataxias. Cerebral white matter lesions, although usually few in number, were observed in significantly more patients than controls, particularly those aged over 50 years.

Torsional nystagmus. A neuro-otological and MRI study of thirty-five cases.

Thirty-five patients with torsional nystagmus (TN) underwent vestibular and ocular motor assessment and magnetic resonance image (MRI) scanning of the head. Patients were divided into two groups according to whether TN was predominant and present in primary gaze (Group I, 23 patients) or elicited by head positioning or gaze deviation and less prominent than other concurrent nystagmus (Group II, 12 patients). The main aetiologies in both groups were demyelination, vascular disease and posterior fossa tumours. In Group I, a frequent pattern of findings, occurring in 30-50% of cases, was a caloric canal paresis contralateral to the direction of the fast phases ('beat') of the TN, whereas the duration of horizontal caloric/rotational nystagmus and the slow-phase eye velocity of pursuit and of optokinetic nystagmus were all reduced in the direction of beating. The TN was more frequently and consistently modulated by vertical canal stimuli (head oscillation in roll) than by otolith stimuli (static tilt). Statistical analysis of the MRI showed significant overlap of abnormal MRI signals in the area of the vestibular nuclei, on the side opposite to the beat direction of TN. These results suggest that TN originates in a central imbalance of vertical semicircular canal function, resulting from lesions involving the vestibular nuclei on the opposite side of the TN. Group II was heterogeneous with no consistent pattern of neuro-otological findings, although lesions ipsilateral to the TN were frequent occurrence; in these cases cerebellar system lesions may have produced ipsilateral vestibular nuclei disinhibition.

The site of brainstem lesions causing semicircular canal paresis: an MRI study.

Ten patients with canal paresis of central origin and ten patients with peripheral canal paresis were studied using MRI of the brainstem to identify lesions within the central vestibular pathways. In the central group, the magnitude of the canal paresis was generally lower than in the peripheral group and removal of fixation had little effect on the nystagmic response. In the peripheral group, removal of fixation enhanced the nystagmus and lessened the discrepancy between the two ears. Statistical processing of the MRI showed that in the central group significant spatially coincident lesions occurred within the medial vestibular nucleus, lateral vestibular nucleus and proximal portion of the vestibular fascicle.

Histopathology of multiple sclerosis lesions detected by magnetic resonance imaging in unfixed postmortem central nervous system tissue.

Postmortem unfixed whole brains from 17 multiple sclerosis and 6 control cases were examined by magnetic resonance imaging (MRI) using a T2-weighted spin echo sequence and histology to determine the relationship between areas of abnormal MRI signal and underlying pathological change. In group 1, small MRI lesions and correspondingly small plaques, most of which were chronic, were detected histologically in 5 brains. In 4 brains there were more extensive areas of both abnormal signal and histological plaques which were more often active (group 2). However, in a further 5 brains extensive MRI abnormalities were observed when only small periventricular plaques were present histologically (group 3). Lesions in the hindbrain and cerebral grey matter were infrequently observed by MRI. The extensive MRI abnormalities seen in areas in which only small histological plaques were found may be the result of vascular permeability changes in the normal-appearing white matter surrounding plaques.

Patterns of blood-brain barrier breakdown in inflammatory demyelination.

The evolution of the changes in the blood-brain barrier (BBB) in chronic relapsing experimental allergic encephalomyelitis (CREAE), a model of immune-mediated demyelination, has been studied by magnetic resonance imaging (MRI); gadolinium-DTPA (Gd-DTPA) was used to detect BBB breakdown by both quantitative and qualitative techniques. Animals with acute EAE were examined for comparison. In animals with CREAE an approximately linear relationship was found between the mean number of lesions enhancing with Gd-DTPA seen per MRI slice and the severity of clinical disability at relapse. In addition, a direct relationship was seen between the duration of clinical relapse and the duration of enhancement with Gd-DTPA for lesions associated with the relapse. Lesions studied in animals having entered a progressive phase of disease showed the most sustained BBB breakdown. These observations suggest that BBB breakdown is important in the development of clinical signs in inflammatory demyelination. In CREAE, areas of focal enhancement with Gd-DTPA could usually be clearly defined at a time of clinical relapse. In slices free of focal lesions, no abnormal Gd-DTPA leakage could be detected using a quantitative method. In contrast, in acute EAE no focal lesions were visible, but significant leakage was detected by measurement. No change was found in T2 relaxation times in CREAE or acute EAE. The pattern of BBB breakdown in inflammatory demyelination evolves from a diffuse shortlived disturbance in acute EAE to a more focal and prolonged breakdown in animals with chronic relapsing and progressive disease. The broad similarities in the pattern of BBB breakdown seen in CREAE and multiple sclerosis support the hypothesis that the initial vascular changes in the human disease are due to inflammation which could be mediated immunologically.

Acute disseminated encephalomyelitis. MRI findings and the distinction from multiple sclerosis.

Brain MRI was performed on 12 patients with acute disseminated encephalomyelitis (ADEM). Multifocal white matter lesions indistinguishable from those seen in multiple sclerosis (MS) were found in 10. In 5 there were rather extensive symmetric abnormalities in the cerebral (2) or cerebellar white matter (2), or basal ganglia (1). Follow-up MRI after intervals of 2 wks to 18 months demonstrated partial resolution of the abnormalities, but some persisting lesions. New MRI lesions were found at follow-up in only 1 case after an interval of 2 wks. Serial MRI makes a useful contribution to the distinction between MS and ADEM.

Duration and selectivity of blood-brain barrier breakdown in chronic relapsing experimental allergic encephalomyelitis studied by gadolinium-DTPA and protein markers.

Gadolinium-DTPA (Gd-DTPA) enhancement seen with magnetic resonance imaging in chronic relapsing experimental allergic encephalomyelitis (CREAE) corresponded with sites of blood-brain barrier breakdown judged by traditional markers in areas of inflammatory demyelination. Duration of Gd-DTPA leakage for individual lesions in CREAE varied from 5 days to more than 5 wks. By contrast, in acute EAE leakage was of shorter duration (always less than 5 days). Selective enhancement was observed in CREAE lesions using Gd-protein markers. Gd-albumin enhancement was not always seen in areas of leakage of the smaller molecular weight compound Gd-DTPA. The addition of immunoglobulin to the gadolinium complex led to enhancement of lesions not seen with Gd-albumin alone. From the similarities between the histology and the patterns of Gd-enhancement in CREAE and multiple sclerosis, it is probable that Gd-enhancement reflects active inflammation (with or without demyelination) in the human disease.

Heterogeneity of blood-brain barrier changes in multiple sclerosis: an MRI study with gadolinium-DTPA enhancement.

We performed 15 dynamic gadolinium-DTPA (Gd-DTPA)-enhanced MRI studies in 8 patients with relapsing and remitting multiple sclerosis; 7 were follow-up studies. We measured the time course of enhancement in 102 enhancing lesions for up to 384 minutes, with rest breaks. Immediate postcontrast MRIs demonstrated many different patterns of enhancement. We observed both uniformly enhancing and ring enhancing lesions. The enhancing regions were often less extensive than the corresponding high signal on T2-weighted images. Three lesions were seen with Gd-DTPA but not on unenhanced scans; 1 was seen on unenhanced scans 10 days later, suggesting that blood-brain barrier disturbance may precede other MRI signs of MS lesions. Three months later, some high-signal areas on T2-weighted scans had decreased in size to resemble the areas previously outlined by Gd-DTPA. This technique provides useful information about the pathogenesis and behavior of MS lesions.

Towards quantitative measurements of relaxation times and other parameters in the brain.

The nature and physical significance of the relaxation times T1 and T2 and of proton density are described. Methods of measuring T1 and T2 are discussed with emphasis on the establishment of precision and the maintenance of accuracy. Reported standards of success are briefly reviewed. We expect sensitivities of the order of 1% to be achievable in serial studies. Although early hopes of disease diagnosis by tissue characterisation were not realised, strict scientific method and careful calibration have made it practicable to apply relaxation time measurement to research into disease process. Serial measurements in patients and correlation with similar studies in animal models, biopsy results and autopsy material taken together have provided new knowledge about cerebral oedema, water compartmentation, alcoholism and the natural history of multiple sclerosis. There are prospects of using measurement to monitor treatment in other diseases with diffuse brain abnormalities invisible on the usual images. Secondarily derived parameters and notably the quantification of blood-brain barrier defect after injection of Gadolinium-DTPA also offer prospects of valuable data.

Pathological findings correlated with magnetic resonance imaging in subcortical arteriosclerotic encephalopathy (Binswanger's disease).

Formalin-fixed brain slices from four cases of subcortical arteriosclerotic encephalopathy in which a firm diagnosis could be made both clinically and pathologically have been studied by magnetic resonance imaging (MRI). The slices were subsequently embedded in paraffin-wax or celloidin and sections were cut in the same plane as the MRI slices. There was a good correlation between the extent and severity of the abnormal MRI signal and the pathological changes. Areas of diffuse MRI abnormality corresponded with areas of axonal and myelin loss with gliosis, and small "lacune"-like lesions corresponded with lacunar infarcts histologically. Sparing of the subcortical U-fibres was seen histologically and on MRI. The abnormal signal probably originates from increased tissue water attributable to gliosis and an expanded extracellular space.

Spasmodic torticollis due to a midbrain lesion in a case of multiple sclerosis.

A case of multiple sclerosis is described in which spasmodic torticollis occurred abruptly and abated after 1 year. Magnetic resonance imaging (MRI) demonstrated a lesion in the mesencephalon. Other symptoms and physical signs that developed at the same time as the spasmodic torticollis were compatible with the lesion that had not been present on MRI 18 months previously. There are very few reports of spasmodic torticollis due to an identified focal lesion; there is evidence from experimental work on animals that midbrain lesions may cause spasmodic torticollis but there has been no previous human example.

Magnetic resonance imaging of the optic nerve in optic neuritis.

Magnetic resonance imaging (MRI) of the optic nerves using the STIR (short inversion time inversion recovery) sequence was performed in 37 adult patients with a recent or past attack of optic neuritis. MRI revealed high-signal regions in 84% of symptomatic and 20% of asymptomatic nerves. The mean longitudinal extent of lesions was 1 cm. Slow or poor visual recovery was associated with more extensive lesions, or lesions within the optic canal. Disk swelling was usually associated with anterior lesions but also occurred with lesions in the canal. Visual evoked potentials were even more sensitive than MRI in detecting lesions and are still the investigation of choice in suspected demyelinating disease involving the optic nerve.

The role of NMR imaging in the assessment of multiple sclerosis and isolated neurological lesions. A quantitative study.

The form and distribution of MRI abnormalities in 114 patients with clinically definite multiple sclerosis (MS) have been compared with observations on 53 apparently healthy individuals, 129 patients with isolated focal neurological lesions with which MS often presents (51 patients with optic neuritis, 44 with isolated brainstem lesions and 34 with isolated spinal cord syndromes) and 105 patients with disorders which may be confused clinically or radiologically with MS. The latter comprised 55 patients with cerebral vascular disease (including 7 cases of dementia with diffuse white matter disease), 24 with degenerative ataxic disorders, 8 with cerebellar tonsillar ectopia, 7 with sarcoidosis and 11 with a variety of other disorders. Periventricular abnormalities were found in all but 2 patients with MS and discrete white matter lesions in all but 12. Characteristically the periventricular changes in MS were irregular in outline. Periventricular abnormalities which were often milder and of smooth outline were seen in 37/55 patients with cerebral vascular disease, 9/24 with cerebellar degeneration, 5/7 with sarcoidosis and in 2/3 apparently healthy individuals over the age of 60. The appearances in the 7 cases of dementia resembled those with advanced MS. Cerebellar and/or brainstem atrophy characteristic of the cerebellar degenerations, in the absence of white matter abnormalities, was helpful in making the distinction from MS. Congenital anomalies and tumours in the region of the brainstem and foramen magnum were readily shown. More than half the patients with symptoms attributable to isolated focal neurological lesions had additional lesions at presentation. MS cannot be diagnosed in these cases at presentation, but repeat scans after 5 to 20 months in 25 patients with optic neuritis and 10 with clinically isolated brainstem lesions have shown new lesions in 7 (20%). The patients with new lesions fulfil the criteria for clinically probable MS (Poser et al., 1983). Measurements of T1 and T2 in vivo permitted the distinction of acute from chronic brainstem lesions. There were quantitative differences in T1 and T2 between the normal appearing white matter in MS and normal brain. Studies of postmortem brains provided convincing evidence that the MRI abnormalities in MS correspond with plaques. Evidence is adduced to support the view that an important source of the abnormal NMR signals in acute lesions is oedema, and in chronic lesions is gliosis; demyelination per se is unlikely to make an important contribution.

STIR sequences in NMR imaging of the optic nerve.

Orbital fat surrounding the optic nerve causes considerable difficulties in NMR imaging due to its high image intensity and the chemical shift artefact. We have investigated the ability of inversion recovery sequences with short inversion times (STIR sequences) to suppress fat signals in imaging the optic nerve. We have also compared the contrast attainable with STIR sequences with that obtainable from other sequences. Measurements were made on 4 normal controls and 5 patients with multiple sclerosis (MS) to obtain typical values of relaxation times and proton densities for orbital fat, cerebral white matter and MS lesions. The fat T1 measurements were used to predict an appropriate inversion time for the STIR sequence and estimate how much residual fat signal might be expected as a result of natural variations in fat T1. STIR sequences can be used to suppress the signal from orbital fat with little residual signal. Measurements from white matter and MS lesions were used to predict the contrast between normal and pathological tissues that is attainable with STIR sequences. STIR contrast compares favourably with that obtainable from other sequences.

MR brain scanning in patients with vasculitis: differentiation from multiple sclerosis.

We performed MR (magnetic resonance) brain imaging on 24 patients with a systemic vasculitis. MRI proved to be a sensitive method for detecting brain lesions (clinically silent or manifest) in these patients. The most frequent abnormalities were periventricular lesions seen in 12 cases. Such changes are not specific for vascular disease, and are often seen in multiple sclerosis. However, additional changes were commonly seen which suggested the correct diagnosis.