Isolated shoulder palsy due to cortical infarction: localisation and electrophysiological correlates of recovery

Abstract

The corticospinal tract influences the distal musculature more than the proximal, and the mechanisms involved in recovery of proximal muscle strength after stroke are unclear. A 65 year old man developed right shoulder weakness due to infarction in the left precentral gyrus. MRI showed a 3 mm cortical–subcortical ischaemic lesion in the superior genu of the left precentral gyrus medially to the knob-like structure corresponding to the motor area of the hand. Two months after stroke, when the patient was able to abduct the right arm against gravity and seven months after stroke when the patient had almost completely recovered, maximal TMS of the contralateral and ipsilateral motor cortex during voluntary contraction did not evoke a MEP in the right deltoid either with a focal or a non-focal coil. Recovery of proximal muscles in these cases may be mediated by elements other than the fast corticospinal neurones responsible for MEP generation.

BACKGROUND

Isolated pure motor involvement of the shoulder and arm muscles is extremely rare after stroke, and up to now has been documented by magnetic resonance imaging (MRI) in only one patient.¹

The corticospinal system is known to exert a greater influence over distal than proximal upper limb muscles, and the mechanisms that induce better recovery from stroke of proximal muscles are debated. The contribution of ipsilateral corticospinal fibres from the unaffected hemisphere, or of corticoreticulospinal projections, has been hypothesised.^2,3

CASE PRESENTATION

A 65-year-old right-handed man awoke unable to abduct the right arm. Examination showed a strength of 0/5 (Medical Research Council scale) in the right deltoid, 2/5 in the biceps and 4/5 in the triceps. Strength in the remaining upper and lower limb muscles as well as the sensory examination were completely normal. Tendon jerks were normal except in the right biceps, where they were reduced. Electromyography showed no voluntary activity in the right deltoid muscle, and a reduced recruitment pattern with motor units of low firing frequency in the biceps brachii. A computed tomography scan showed a small hypodense lesion in the left precentral gyrus. Two months later, the patient was able to abduct the arm against gravity and the strength in the biceps was 4/5. Biceps tendon jerk was hyperactive. At 7 months, the patient was able to hold the right arm abducted against resistance (4+/5)

INVESTIGATIONS

Standard and functional magnetic resonance imaging (fMRI) was performed 2 months after stroke. fMRI was carried out using the blood oxygenation level-dependent contrast technique. To explore the cortical activity of the primary and secondary motor cortices, a self-paced finger-tapping task at a frequency of about 1 Hz was used. The experimental paradigm was a block design alternating a task condition (right or left finger tapping) with a control state having the same duration (30 s; 3 s for each fMRI volume) and consisting of complete motor and mental relaxation. For each run, 135 volumes were acquired. Raw data were analysed by means of the Brain Voyager 4.9 software (Brain Innovation, Maastricht, The Netherlands) and statistical analysis was performed using the general linear model. Statistical maps were thresholded at p<0.001 at the voxel level and a cluster size of at least four voxels was required. Functional volumes were co registered with a high-resolution structural TSE T2-weighted sequence where the ischaemic lesion was detectable. MRI showed a 3 mm cortical–subcortical ischaemic lesion in the superior genu of the left precentral gyrus medially to the knob-like structure corresponding to the motor area of the hand activated by the right finger-tapping task (fig 1A).

Figure 1. (A) Axial T2-weighted image with superimposed functional magnetic resonance imaging activations obtained during a right finger-tapping task.

Standard T2-weighted sequence shows an ischaemic lesion (arrow) in the superior genu of the left precentral gyrus medially to the structure activated by the finger tapping. An activated region was also present in the mesial surface of the left frontal lobe corresponding to the supplementary motor area. Image is shown using the right–left radiological convention. (B) Cortical representation maps of the deltoid, biceps and first dorsal interosseus (FDI) muscles 2 months after stroke. The stimulus intensity was set at 200% of resting motor threshold, defined as the minimum stimulus intensity that produced a liminal motor-evoked potential (MEP) (about 50 μV in 50% of 10 trials) at rest in the FDI muscle. At this intensity, a MEP of about 3 mV was evoked in the relaxed FDI and of about 0.5 mV in the proximal muscles when stimulating the area around the optimal scalp position for each muscle. The responses evoked by three stimuli at each scalp site were averaged. The area of each map was taken as the number of scalp points at which a response was evoked, and the volume of each map was taken as the sum of the averaged MEPs recorded at all scalp sites at which a response was evoked. Each point of the scalp surface is identified by two coordinates on an x–y plane; the x axis lies on the frontal plane and the y axis lies on the longitudinal plane. The white cross indicates Cz. Positive and negative points are located on the right and left hemisphere, respectively, on the x-axis, and anterior and posterior to the line of Cz on the y axis. Cortical maps represent the motor-evoked potentials after stimulation of each point of the scalp recorded on the contralateral muscle. The amplitude of the response on each point is represented on the z axis. There are no MEPs in the right deltoid.

Magnetic stimulation was performed two and seven months after stroke using a high-power Magstim 200 magnetic stimulator (Magstim Co., Whitland, Dyfed, UK) connected to a figure-of-eight coil. Motor-evoked potentials (MEPs) were recorded bilaterally and simultaneously from the deltoid, biceps brachii and FDI muscles. To investigate the presence of contralateral and ipsilateral responses, we stimulated the optimal scalp position for each of the three muscles studied in the affected and unaffected hemisphere using a stimulus intensity of 100% of maximum stimulator output during tonic activation at about 20% of maximum voluntary contraction. To increase the possibility of detecting MEPs in paretic muscles, cortical stimulation was also performed with a non-focal circular coil centred over the vertex; moreover, the scalp was mapped systematically using a figure-of-eight coil and a standard TMS protocol evaluating the cortical representation of each tested muscle together with the deltoid muscle.⁴ Cortical stimulation with a non-focal circular coil centred over the vertex was performed at an intensity corresponding to the maximum stimulator output, with a clockwise inducing current flow as viewed from above for the right motor cortex and vice versa for the left motor cortex, both at rest and during voluntary contraction of the tested muscles at approximately 20% of the maximum.

Two months after stroke, maximal TMS of the contralateral and ipsilateral motor cortex during voluntary contraction did not evoke a MEP in the right deltoid either with a focal or a non-focal coil. TMS mapping showed no representation of the right deltoid in the left motor cortex and adjacent areas (fig 1B). A large-amplitude MEP was recorded at rest in the left deltoid after stimulation of the opposite motor cortex (fig 1B); this might be due to an increase in the excitability of the right, intact motor cortex consequent to an imbalance in interhemispheric inhibition.⁵ Seven months after stroke, when the patient had almost completely recovered, maximal TMS of the contralateral and ipsilateral motor cortex still did to not evoke an MEP in the right deltoid either with a focal or a non-focal coil.

DISCUSSION

Recently, fMRI studies allowed us to correlate the somatotopy of the primary motor cortex to anatomical landmarks. A study indicated that the segment of the precentral gyrus containing the hand function is a knob-like structure corresponding to the middle knee of the central sulcus shaped like an inverted omega.⁶ It has been difficult to obtain a “pure” fMRI cortical activation of the shoulder muscles because the movement also propagates to the distal muscles during a shoulder motor task. Crafton et al⁷ demonstrated that 63% of areas activated during shoulder movement were also activated during hand motor tasks, but they were able to individuate a centre of activation for the shoulder with Talairach’s coordinates 23, –29 and 55. These coordinates corresponded with the area where the ischaemic lesion was localised in the patient we report. This area was medial to the hand cortical activation area obtained by a finger-tapping task as expected from the classical human primary motor cortex map.

The patient we report recovered almost completely, but we could not record MEPs in the right deltoid muscle after maximal stimulation of the contralateral motor cortex, and there was no representation of the right deltoid in the left motor cortex and adjacent areas. MEPs were not evoked also by maximal stimulation of the ipsilateral motor cortex. Few other patients who recovered voluntary activity in proximal muscles but had no motor responses to contralateral and ipsilateral TMS have been reported.² Recovery of proximal muscles in these cases may be mediated by elements other than the fast corticospinal neurones responsible for MEP generation. Slowly conducting contralateral or ipsilateral corticospinal projections or polysynaptic pathways, such as corticobulbospinal, corticoreticulospinal or corticopropriospinal projections, which cannot be probed by TMS, may compensate the reduction or even the complete loss of the fast corticospinal inputs to the motoneurones of proximal muscles after stroke.

LEARNING POINTS

• The corticospinal tract is known to be important for recovery of distal muscle function after stroke.

• The mechanisms responsible for recovery of proximal muscle power after stroke are unclear.

• In the patient reported here recovery of proximal muscle power was not mediated by fast corticospinal neurons in the contralateral and ipsilateral motor cortex but by slowly conducting contralateral or ipsilateral corticospinal projections or polysynaptic pathways which cannot be probed by TMS.