, Volume 45, Issue 10, pp 708–712

Corticobasal degeneration: structural and functional MRI and single-photon emission computed tomography


    • Department of RadiologyOspedale di Cattinara, University of Trieste
  • R. Moretti
    • Department of Physiology and PathologyOspedale di Cattinara, University of Trieste
  • P. Torre
    • Department of Internal Medicine and Clinical NeurologyOspedale di Cattinara, University of Trieste
  • R. M. Antonello
    • Department of Internal Medicine and Clinical NeurologyOspedale di Cattinara, University of Trieste
  • R. Longo
    • Department of PhysicsUniversity of Trieste
  • A. Bava
    • Department of Physiology and PathologyOspedale di Cattinara, University of Trieste
Diagnostic Neuroradiology

DOI: 10.1007/s00234-003-1058-1

Cite this article as:
Ukmar, M., Moretti, R., Torre, P. et al. Neuroradiology (2003) 45: 708. doi:10.1007/s00234-003-1058-1


We studied seven patients with corticobasal degeneration (CBD) from a clinical and imaging perspective. We describe the main morphological features of CBD and, using functional MRI, try to define the possible role of the parietal lobe in simple and complex learned motor sequences. We showed decreased activation of the parietal lobe contralateral to the more affected arm, when movements, simple or complex, are performed with that hand. Moreover we found that functional imaging can demonstrate parietal and motor cortex dysfunction before structural, and even single-photon emission computed tomography changes become evident.


Corticobasal degenerationParietal lobeFunctional magnetic resonance imagingSingle-photon emission computed tomography


Corticobasal degeneration (CBD) most frequently causes asymmetrical parkinsonism with cortical abnormalities. Two sizeable series including 51 patients, five with pathological verification, have been reported, and their signs and symptoms discussed [1, 2]. The most common were limb rigidity (100%), limb apraxia (91%), gait difficulties (89%), focal reflex myoclonus (88%), eye movement abnormalities (78%), limb dystonia (77%), pyramidal signs (73%), dysarthria (62%), cortical sensory abnormalities (55%) and the alien limb phenomenon (55%). Tremor, frontal lobe features, and cognitive impairment were less common. However, a recent review of Canadian Brain Tissue Bank cases [3, 4] suggests that dementia may be the commonest presentation, rather than the better recognised perceptual-motor syndrome previously described. The disease is steadily progressive, with a mean survival of 6–7 years.

Neuropathologically, circumscribed parietal or frontoparietal lobar atrophy may be present [5]. Severe neuronal loss and intense astrogliosis are evident in the cortex; spongiosis, swollen and achromatic neurones (ballooned cells or pale bodies), neuropil threads and neurofibrillary tangles can be found. Ballooned cells can be detected in the neocortex; basophilic argyrophilia and tau-positive inclusions are found in the neurones of the substantia nigra and subthalamic nucleus, striatum and pallidum and even along the dentate-rubro-thalamic tracts [6, 7]. Astrocytic plaques have been described as characteristic [8].

However, while most cases of CBD show the characteristic ballooned neurones, tau-positive neuronal and glial inclusions are not infrequent, and some other cases are difficult to classify due to the presence of overlapping neuropathological features of Alzheimer's disease, progressive supranuclear palsy, Parkinson's disease and pure hippocampal sclerosis [9].

A prominent feature in almost all cases of CBD is apraxia, a disturbance of goal-directed motor behaviour characterised by inability to perform previously learned movements in the absence of weakness or sensory defects. There is a striking preservation of perception, attention, coordination, motivation and comprehension: thus it is a high-level disorder of movement representation. A study on the less involved limb of ten patients with CBD showed that seven had ideomotor and none buccofacial apraxia [10].

There is little information on movement organization in CBD. We therefore studied seven patients with CBD, to examine the disruption of fine distal movement execution, not only clinically, but also by using functional MRI (fMRI).

Materials and methods

We studied two men and five women complaining mainly of an asymmetrical akinetic syndrome, affecting the left arm in all cases. Their mean age was 66.6 years (±7.6 years). All were right-handed (+22.6 average score on the Briggs and Nebes test) [11]. None gave a history of cerebrovascular disease, hypertension or metabolic disorders. Their most common complaint was a relatively recent asymmetrical akinetic syndrome, bradykinesia, dystonic postures, alien limb syndrome (in two), slurred speech and gait difficulty; two had action tremor. All underwent a complete examination of eye movements, showing normal saccadic velocity (considering antisaccades, reflexive saccades and voluntary saccades), with increased latency (especially of voluntary saccades) and preserved pursuit and optokinetic nystagmus. One showed a supranuclear palsy, affecting vertical and horizontal gaze. None had a positive response to levo-dopa, and no autonomic disturbances were detected. The mean duration of symptoms was 8.4±2.8 months at the time the patients were studied.

Intelligence performance was tested using the Raven standard progressive matrices [12]; right/left personal and extrapersonal hemispace recognition, and any signs of tactile agnosia or buccofacial apraxia were assessed by the tests of Spinnler and Tognoni [13]. Global cognition was tested using the Wechsler adult intelligence scale (WAIS) [14], linguistic fluency and appropriateness using the Goodglass and Kaplan scale [15] and memory performance using the Wechsler memory scale (WMS) [16]. Visuospatial perception was assessed with the Benton line orientation test [17] and executive functions using the Clock test [18].

All patients underwent MRI at 1.5 T, with axial and coronal slices. We used the following sequences: proton density and T2 weighted spin-echo (SE), T2-weighted fast SE and fast fluid-attenuated inversion-recovery.

All patients also underwent fMRI. During the observation period, they were invited to execute three or four training sessions daily, to practice a finger-opposition task, starting with the left (affected) hand and ending with the right. They had to touch all four fingers with the thumb, beginning with the index finger proceeding to the small finger, then repeating this sequence. Then they were asked to execute a complex alternating sequence, starting with the left hand and ending with the right. They had to touch the fingers in a different sequence: 1-2, 1-4, 1-3 and 1-5. The acquisition time was equally divided into three motor task periods, alternating with three rest periods. We collected seven axial images per period, so in each study 42 images were acquired. The subject's head was immobilised with rubber foam wedges between the head and the support. Field homogeneity was adjusted by iterative automatic shimming in each examination. The gradient-recalled (GRE) images were obtained on a standard 1.5 T imager, maximum gradient strength was 10 mT/m. A standard quadrature birdcage head coil was used as the receiver coil, and the body coil for excitation.

The major parameters of the 2D gradient-echo pulse sequence were TR 60 TE 40 ms, flip angle 25 degrees, field of view 160×144 mm, slice thickness 4 mm, matrix 128×128. The T1 contrast enhancement option was activated [19], and an MRA acquisition was performed at each T1-GRE fMRI acquisition. The major parameters of the angiographic sequence were TR shortest, flip angle 20 degrees, field of view as for T1 GRE, slice thickness 1 mm, matrix=256×256, 12 slices, phase contrast technique.

Image analysis was performed using a program in IDL Environment (Interactive Data Language, Research System Inc., USA). The basic analysis consists of the calculation of the correlation coefficient between time-intensity behaviour of each pixel and a square-wave model function. To exclude transient haemodynamic responses, five images per block (the 3rd to the 7th) are included in the analysis. By applying a correlation analysis ( P <0.001) and cluster filtering (at least five pixels) a raw activation map was obtained.

This map is affected by flow artefacts, to eliminate which the activation map and MRA were compared and activation clusters related to vessels were rejected. An fMRI data-analysis procedure based on an innovative approach to motion artefact reduction, in which no image registration procedures are required, was applied. Stimulus-correlated artefacts were rejected by comparison between the raw activation map and the intensity gradient image [20]. In the final step of data analysis the clusters are expanded until their statistical significance remains constant and highly significant [20]. In the figures the original clusters are white, and the black pixels are due to the expansion procedure. The region of activation was then superimposed on the anatomical image to produce individual activation maps.

During the rest phases, the subjects performed no voluntary motor activity. During the task phases, they repeatedly opposed the thumb to the other fingers, as described. The task was done in a self-placed manner, and the right or left hand was used on the verbal command of the examiner. All subjects manifested clear problems in executing the motor sequence with the left (affected) hand; in order to achieve a better result, they tried to control their task with a constant visual effort. Although that is clearly impossible during fMRI, the patients nevertheless continually attempted this in the imager, so that the fMRI images suffered movement artefact.

All patients also underwent 99-mTc-ECD single-photon emission computed tomography (SPECT), with acquisition every 35 s for a total of 64 images. This was done at least 1 week after the beginning of the study.


Intelligence performance was in the normal range (106±4.5) by Raven standard progressive matrices; the patients recognised right/left personal and extrapersonal hemispace, and showed no sign of tactile agnosia or buccofacial apraxia. None could pantomime to verbal command, whereas they could to some extent imitate pantomimes performed by the examiner: all showed signs of ideomotor apraxia and none showed ideational apraxia, with the left, affected hand.

WAIS average results demonstrated a mild general tendency to overall deterioration (12.4+4.7%). The patients obtained an average score of 3.4+0.76 on the Goodglass and Kaplan Scale. WMS showed an average score of 67.45+3.24), reflecting mild deterioration of logical, procedural and verbal memory strategies. The Benton line orientation test demonstrated a mild tendency to ignore left hemispace and clumsy hand movements. All patients performed poorly on the Clock test (mean score 5.6+1.76).

Structural MRI was normal in one patient, but in the other six there was asymmetrical perirolandic cortical atrophy, mild atrophy of the basal ganglia and slightly increased signal in pre- and postrolandic cortex (Fig. 1). In one patient, there was also low signal from the lentiform nuclei.
Fig. 1 A, B.

Structural MRI. A T2-weighted spin-echo image showing mild asymmetrical atrophy of the right prerolandic cortex. In the right pre- and postrolandic cortex signal is slightly increased; this is shown better in B, a fast fluid-attenuated inversion-recovery image. (In all figures, the images are of single patient, but the findings were common to all)

On fMRI, during the execution of a simple motor task with the right, healthy hand, there was a good activation of the left rolandic cortex, supplementary motor (SMA) and left parietal cortex, and of the right prefrontal cortex. In contrast, during performance of the same task with the left hand, there was reduced activation of the right rolandic and parietal cortex and SMA (Fig. 2).
Fig. 2 A, B.

Functional MRI (fMRI), with a simple motor task: A Right hand movement: good activation of the left rolandic cortex, supplementary motor area (SMA) and parietal cortex, and the right prefrontal cortex. B Left, affected, hand movement: decreased activation of the right rolandic cortex, SMA and parietal cortex

During performance of the complex motor task with the right hand there was bilateral activation of rolandic and parietal cortex, and of the left frontal cortex. During performance of the same task with the left hand there was bilateral activation of rolandic cortex and SMA, but only very modest bilateral activation of the parietal cortex, more evident on the right (Fig. 3).
Fig. 3 A, B.

f-MRI, with a complex motor task. A Right hand movement: bilateral activation of rolandic and parietal cortex, and of left frontal cortex. B Left, affected, hand movement: bilateral activation of rolandic cortex and SMA, but very modest bilateral activation of parietal cortex, more evident on the right

In all the patients examined SPECT showed slight right posterior parietal hypoperfusion, with concomitant mildly reduced perfusion of the right temporal and frontal cortex. A lesser degree of inhomogeneous hypoperfusion was seen in the left parietal cortex.


We have focused our attention on the results obtained with fMRI as far as motor cortical organisation is concerned. Since there are no laboratory markers for CBD, neuropathology remains the gold standard. MRI can support the diagnosis when it shows asymmetrical ventricular enlargement and enlargement of parietal sulci [21]. SPECT can be helpful when it shows asymmetrical parietal hypoperfusion [22].

However, MRI and SPECT are not pathognomonic, in particular in the early stages. In contrast, even in early stages of CBD (as reported here), fMRI can provide indirect evidence of asymmetrical disorganisation of the hierarchical cortical motor programme, mainly due to primary parietal lobe dysfunction.

Gesture disorders of parietal lesions result in a lack of manual dexterity, reminiscent of kinaesthetic limb apraxia. However, gesture identification is preserved, in contrast to the involvement of all aspects of gesture execution, suggesting that the mental representation or conceptual aspects of gestures are not involved.

A motor sequence (such as finger opposition) is an example of movements in intrapersonal space: the fingers of the hand are moved in relation to the thumb. The body reference coordinate system is automatically updated. Apart from the obvious activation of primary somatosensory and supplementary sensory area, movements conducted in intrapersonal space activate the parietal lobe [23]. Moreover, there are important circuits between the inferior parietal lobe and inferior premotor area, which are not simply part of a motor system, activated by sensory stimuli and eventually triggering actions on the basis of this information. Their functions are much more complex and consist of storing elementary motor programmes and retrieving these to interact with the environment [24].

Moreover, the parietal lobe participates in motor control, especially as far as spatial control and coordinate-transformation system for sensory driven strategies of the eyes, arm and hand is concerned, through its intimate connections with the cerebellum [25]. Our holistic impression of space, as well as awareness of an internal intention to make a movement, may be coded in this abstract, distributed representation of space in the posterior parietal cortex [26, 27].

The clinical impression of all our patients was is that execution, which normally derives from internal planning and from the storage of knowledge, is disrupted. Our evidence of right parietal dysfunction leads to the conclusion that parietal dysfunction, by reducing the interconnections between the SMA, premotor and motor areas, disrupts the storage, retrieval, and effective execution, of complex learned movement.

fMRI helps to reveal in vivo the neuroanatomical substrate of CBD; it also provides evidence of altered higher cortical motor organisation early in the disease, often before structural changes of parietal atrophy are evident.


This study was supported by grants from the Ministero dell'Università e della Ricerca Scientifica e tecnologica, Roma, from the Università degli Studi di Trieste and from the Consorzio per lo Sviluppo Internazionale dell'Università degli Studi di Trieste and by Fondazione C. e D. Callerio, Laboratori di Ricerche Biologiche, Trieste, Italia.

Copyright information

© Springer-Verlag 2003