Multiple System Atrophy

The Parkinson's-Reversing Breakthrough

Parkinson Disease Guide

Get Instant Access

MSA has been separated into MSA-P and MSA-C according to the predominant parkinsonian or cerebellar features. The former corresponds to the old term striatonigral degeneration (SND), the latter to sporadic olivopontocerebellar atrophy (OPCA). The term Shy-Drager syndrome is no longer been considered useful by the Consensus Conference on MSA (11), but a proposal to introduce the term MSA-A (autonomic features-predominant MSA) was recently made by Horimoto et al. (12) to indicate those MSA patients in whom autonomic dysfunction predominated over cerebellar or par-kinsonian features.

For the purpose of identifying the MRI characteristic features or markers or pointers of the different atypical parkinsonian disorders, one should consider the pathology of the specific disorder (13). There is no doubt that in MSA-P or SND the pathology is mostly in the putamen, whereas in MSA-C the changes are in the brainstem and cerebellum. In MSA-A, most of the abnormalities are in the spinal cord and may be, therefore, disregarded when considering MRI.

Multiple System Atrophy of the Parkinsonian Type

The MRI abnormalities observed in MSA-P have been described in several papers; they should be considered separately according to whether they are obtained at low- or high-field intensity MRI (14-19).

Low-field intensity MRI (up to 0.5T) demonstrates increased signal intensity in proton density and T2-weighted images in the dorsolateral and posterior part of the putamen, where loss of neurons and gliosis in a loose tissue have been described (20,21). These findings correspond to an increased amount of water in the tissue (Fig. 1). Atrophy of the putamen has been commonly observed (22) but only recently emphasized by Yekhlef et al. (23).

High-field intensity MRI (usually 1.5T) shows a different feature, i.e., putaminal hypointensity in T2-weighted images that may completely or almost completely mask the hyperintensity described above (16-19,24-26). The hypointensity is caused by a magnetic susceptibility effect mainly owing to deposits of iron in the same putaminal areas (27); this effect is proportional to the square of the magnetic field intensity and is, therefore, usually evident only at high-field MRI. A lateral rim of hyperintensity, indicating a band devoid of iron right on the lateral margin of the posterior part of the putamen, may be recognizable (25), particularly if we observe proton density and T2-weighted images together. In fact, the magnetic susceptibility effect is much less evident in proton density images. This "slit-like" lateral rim of hyperintensity has been briefly called the "putaminal slit" (12). In conclusion, at 1.5 T MRI putaminal hypointensity in T2-weighted images more marked in the posterior part of this nucleus and the associated thin lateral rim of hyperintensity are the features of MSA-P (17,25,26,28,29) (Fig. 2).

Slit Like Rim Msa Image
Fig.1. MSA-P. 0.5 T, coronal T2-weighted image shows increased signal intensity in the right putamen in a patient with parkinsonism prevalent on the left side.

A very thin line of milder hyperintensity along the entire lateral profile of the putamen, at the interface with the external capsule, may be confused with the "lateral hyperintense rim" or "putaminal slit." This very subtle line has been occasionally observed in atypical parkinsonian disorders (30). In our experience, a very thin line that also borders the anterior part of the putamen, often without associated hypointensity, may be a nonsignificant finding since we have observed it, although rarely, mainly in normal subjects (Fig. 2B). As in the figure shown by Macia et al. (30), it should be noted that the evidence of this line is mostly along the anterior part of the putamen, whereas the abnormalities documented by pathology in MSA-P and shown by MRI typically affect the posterior putamen (21,27).

All these details are important when one is investigating sensitivity and specificity of MRI abnormalities in MSA-P. Specificity, sensitivity, and positive predictive value of the putaminal abnormalities have been investigated in a limited number of studies (18,19,23). The specificity is high, but sensitivity has been considered low, generally not higher than 50-60% (18,19). As most neuroradiologists know, conventional spin-echo (CSE) images are more sensitive than fast spin-echo (FSE) images to magnetic susceptibility effects owing to iron deposits in the tissue. In MSA-P, Righini et al. (31) have elegantly demonstrated that, by using CSE thin sections (3 mm) and comparison of proton density and T2-weighted images vs FSE 5 mm T2-weighted images, the sensitivity is increased from 45% to more than 83%.

Another trivial explanation of the difference in sensitivity between different series is the length of the disease at time of MRI examination. Only very few investigations have dealt with the MRI demonstration of the progression of MSA (12,26). In the series reported by Watanabe et al. (26), frequency of putaminal abnormalities went up from 38.5% in patients with 2 yr or less from motor impairment to 80% in patients with more than 4 yr from motor impairment.

Mri Multiple Systems Atrophy

Of course, MRI abnormalities supporting the diagnosis of MSA-P have a greater importance if they are recognizable in the early stages of the disease, when the clinical diagnosis is still uncertain, rather than in its late stages. Detection of subtle abnormalities that may indicate the diagnosis of MSA-P rather than PD on the first MRI examination, performed at the time of the first visit by a neurologist, may occur, but such episodes are still anecdotical. How early the abnormalities appear is an aspect that needs to be clarified.

A few proton MR spectroscopy (MRS) studies have demonstrated that N-acetylaspartate (NAA) and choline (Cho) were reduced in the lentiform nucleus of MSA patients (32,33). The NAA/creatine (Cr) ratio was particularly decreased in MSA-P patients, probably reflecting putaminal neuronal loss (32). MRS, however, has not been widely used to help differentiate patients with MSA from patients with PD.

The use of diffusion-weighted imaging (DWI) has been recently proposed to help differentiating MSA-P from PD (34). The Innsbruck group demonstrated that the patients with MSA-P presented an increased regional apparent diffusion coefficient (rADC) in the putamen compared with both patients with PD and normal subjects (34). In another study, however, DWI failed to discriminate MSA-P and PSP (35). The advantage of DWI is that one can obtain measurements and quantification. It is not clear, however, whether DWI studies of the basal ganglia in MSA-P patients will prove to have a greater sensitivity than the CSE sequences of MRI, as described by Righini et al. (31).

Multiple System Atrophy of the Cerebellar Type

In MSA-C there are two types of abnormalities, atrophy and signal changes, which do not depend on magnetic field intensity (36). Atrophy has a very characteristic distribution; it mainly involves the pons, middle cerebellar peduncles, and cerebellum. Therefore, the diagnosis can be suspected when one simply observes, in the MRI sagittal midline section, the decreased bulking of the pons compared to the midbrain and particularly the medulla oblongata. The flattening of the profile of the pons begins, in fact, in its caudal part (Fig. 3A). In addition to atrophy, signal changes may be recognized in proton density and T2-weighted images consisting of slight hyperintensity in the structures that degenerate or become gliotic (36). They include: (a) the pontine nuclei and their fibers (the transverse pontine fibers) that run mostly on the anterior and posterior aspect of the basis pontis, cross the midline on the raphe, and reach the cerebellum through the middle cerebellar peduncles (Figs. 3B and C); and (b) the whole cerebellum, because of loss of Purkinje cells and degeneration of their fibers and consequent diffuse gliosis. To make the resulting MRI picture more characteristic, these signal abnormalities are combined with the signal preservation of the structures that do not degenerate (17). They are mainly the superior cerebellar peduncles and the corticospinal tracts. The axial sections of the pons, therefore, present a typical aspect with a sort of cross recognized in the early papers (36), later defined as the "hot cross bun sign" (18,26) or simply the "cross sign" (12) (Figs. 3B and C). Coronal sections are particularly useful to demonstrate the cerebellar hyperintensity when it is very slight, by allowing a comparison in the same image of the infratentorial and the supratentorial structures (17,36). They also beautifully demonstrate the atrophy of the middle cerebellar peduncles (36).

Fig. 2. (opposite page) MSA-P. 1.5 T, axial sections (A-E). First examination (A,B) performed when the patient, with a 2-yr history of dysautonomia, developed mild right-sided parkinsonism. The proton density image (A) is normal whereas the T2-weighted image (B) shows mildly decreased signal intensity in the posterior part of the left putamen (curved arrow), consistent with iron deposits. The very subtle bands of hyperintensity along the anterior part of the putamina (arrowheads) probably are a nonsignificant finding. On the follow-up examination performed 19 mo later, when the parkinsonian signs always prevalent on the right side had progressed, hyperintensity in proton density image (C) in the left posterior putamen has developed (arrowheads). In the T2-weighted sections (D and, adjacent cranial section, E), more marked hypointensity and a lateral rim of hyperintensity are visible.

Multiple Systems Atrophy Cerebellar

Fig. 3. MSA-C. T1-weighted midline sagittal section in an early case (A) shows flattening of the inferior part of the profile of the pons and cerebellar atrophy. In an advanced case, T2-weighted axial sections on the lower (B) and upper pons (C) show atrophy of the pons, middle cerebellar peduncles, and cerebellum with characteristic signal changes. The left corticospinal tract is indicated by an arrow; the margins of the hyperintense middle cerebellar peduncles (open arrows) are poorly defined with respect to the CSF (B). (C) Arrowheads point to the hyperintense anterior and posterior right transverse pontine fibers; an arrow indicates the left superior cerebellar peduncle. On the midline, the raphe is hyperintense.

Fig. 3. MSA-C. T1-weighted midline sagittal section in an early case (A) shows flattening of the inferior part of the profile of the pons and cerebellar atrophy. In an advanced case, T2-weighted axial sections on the lower (B) and upper pons (C) show atrophy of the pons, middle cerebellar peduncles, and cerebellum with characteristic signal changes. The left corticospinal tract is indicated by an arrow; the margins of the hyperintense middle cerebellar peduncles (open arrows) are poorly defined with respect to the CSF (B). (C) Arrowheads point to the hyperintense anterior and posterior right transverse pontine fibers; an arrow indicates the left superior cerebellar peduncle. On the midline, the raphe is hyperintense.

Both specificity and sensitivity of MRI pontine and cerebellar signs in MSA-C are very high (13,18,19). In the series reviewed by Schrag et al. (19), 83% of MSA-C patients could be unequivocally classified based on the MRI findings. If the disease is advanced, sensitivity may be greater than 90% (26). Specificity may decrease if we do not exclude patients with hereditary ataxias (25). Patients with spinocerebellar ataxias (SCAs), particularly SCA1, SCA2, and SCA3, may share the same features both at MRI and histology (37-39). A few of our cases of SCA2, which together with SCA1 is the most common inherited spinocerebellar ataxia in Italy, had an MRI scan that was indistinguishable from that of a patient with MSA-C except for the presence of some degree of cerebral atrophy.

As with MSA-P, one problem in MSA-C is to know how early the MRI signs of atrophy and signal changes develop in the course of the disease, i.e., how helpful can MRI be in the early stages of MSA-C. According to Horimoto et al. (12), the initial changes of the "cross sign" were seen in all their cases of MSA-C within the first 2 or 3 yr after onset of symptoms. As expected, patients with MSA-C normally present pontine and cerebellar changes earlier, with greater severity and more rapid progression than other signs of MSA such as putaminal changes, which usually appear later and have a slower progression. Reciprocally, patients with MSA-P have earlier and more severe putaminal involvement, though posterior fossa changes in these patients may appear later and remain milder. It is, therefore, somewhat misleading to consider indifferently the putaminal and the posterior fossa abnormalities as indicators of MSA and to verify their specificity and sensitivity. In our opinion, one must always consider the clinical presentation (MSA-P or MSA-C or perhaps MSA-A) and should then investigate whether the appropriate MRI signs, i.e., those fitting with the diagnosis, are present (13,17,19,23,26). Occasionally, however, the "inappropriate" MRI sign had an earlier progression than the clinically appropriate one (12). Of course, both putaminal and brainstem-cerebellar signs should be present in a "full-blown" MSA.

Large multicentric studies including hundreds of patients from different countries will probably answer all the questions regarding specificity, sensitivity, positive predictive values, time of appearance, and progression of the MRI signs in MSA-P, MSA-C, and PSP (40).

In the attempt to find measurable data, Kanazawa et al. (41) investigated brainstem and cerebellar involvement in patients with MSA-C using DWI. They found significant differences in the apparent diffusion coefficient (ADC) values of the pons, middle cerebellar peduncles, cerebellar white matter, and putamen between MSA-C patients and controls. The ADC values also correlated with the duration of the disease. ADC values of MSA-C patients in the pons, middle cerebellar peduncles, and cerebellar white matter were also higher than those obtained in a few patients with SCA3 and SCA6.

Measurable data can also be obtained by MRS. Mascalchi et al. demonstrated decreased NAA/Cr and Cho/Cr ratios in the pons and cerebellum of MSA-C patients vs controls (42). MRS, however, is a complex and time-consuming examination that is performed in only a few specialized centers. To obtain measurable data, it is easier to obtain DWI sequences that require a very short acquisition time and calculate the ADC maps (35).

Measurable data are required when dealing with large number of patients in research studies. They are also necessary in longitudinal studies to quantify the progression of the disease. They are hardly needed in the everyday examinations of patients with atypical parkinsonian disorders in whom a diagnosis or a support to a specific diagnosis is required by the clinician. Specificity and sensitivity of MRI findings in these disorders are not yet fully established; much more time will be needed to evaluate sensitivity and specificity in diffusion or spectroscopy studies.

Was this article helpful?

+2 0
Unraveling Alzheimers Disease

Unraveling Alzheimers Disease

I leave absolutely nothing out! Everything that I learned about Alzheimer’s I share with you. This is the most comprehensive report on Alzheimer’s you will ever read. No stone is left unturned in this comprehensive report.

Get My Free Ebook


Post a comment