Anatomic and Physiologic Considerations
The terms pyramidal, corticospinal, and upper motor neuron are often used interchangeably, although they are not altogether synonymous. The pyramidal tract, strictly speaking, designates only those fibers that course longitudinally in the pyramid of the medulla oblongata. Of all the fiber bundles in the brain, the pyramidal tract has been known for the longest time, the first accurate description having been given by Türck in 1851. It descends from the cerebral cortex; traverses the subcortical white matter (corona radiata), internal capsule, cerebral peduncle, basis pontis (ventral pons), and pyramid of the upper medulla; decussates in the lower medulla; and continues its caudal course in the lateral funiculus (column) of the spinal cord—hence the alternative name corticospinal tract. This is the only direct long-fiber connection between the cerebral cortex and the spinal cord (Fig. 3-2). The indirect pathways through which the cortex influences spinal motor neurons are the rubrospinal, reticulospinal, vestibulospinal, and tectospinal; these tracts do not run in the pyramid. All of these pathways, direct and indirect, are embraced by the term upper motor neuron or supranuclear, meaning above the anterior horn cells.
Corticospinal and corticobulbar tracts. The various lines indicate the trajectories of these pathways, from their origin in particular parts of the cerebral cortex to their nuclei of termination.
A major source of confusion about the pyramidal tract stems from the traditional view, formulated at the turn of the 20th century, that it originates entirely from the large motor cells of Betz in the fifth layer of the precentral convolution (the primary motor cortex, or area 4 of Brodmann1) (Figs. 3-3 and 22-1). However, there are only some 25,000 to 35,000 Betz cells, whereas the medullary pyramid contains about 1 million axons (Lassek). Thus the pyramidal tract contains many fibers that arise from cortical neurons other than Betz cells, particularly in Brodmann areas 4 and 6 (the frontal cortex immediately rostral to area 4, including the medial portion of the superior frontal gyrus, i.e., the supplementary motor area); in the primary somatosensory cortex (Brodmann areas 3, 1, and 2); and in the superior parietal lobule (areas 5 and 7). Data concerning the origin of the pyramidal tract in humans are scanty, but in the monkey, by counting the pyramidal axons that remained after cortical excisions and long survival periods, Russell and DeMyer found that 40 percent of the descending axons arose in the parietal lobe, 31 percent in motor area 4, and the remaining 29 percent in premotor area 6. Studies of retrograde transport of tracer substance in the monkey have confirmed these findings.
Lateral (A) and medial (B) surfaces of the human cerebral hemispheres, showing the areas of excitable cortex, i.e., areas, numbered according to the scheme of Brodmann. (Reprinted with permission from House EL, Pansky B: A Functional Approach to Neuroanatomy, 2nd ed. New York, McGraw-Hill, 1967.) See also Fig. 22-1.
Fibers from the motor and premotor cortices (Brodmann areas 4 and 6, Fig. 22-1), supplementary motor cortex, and portions of parietal cortex (areas 1, 3, 5, and 7) converge in the corona radiata and descend through the posterior limb of the internal capsule, basis pedunculi, basis pontis, and medulla. As the corticospinal tracts descend in the cerebrum and brainstem, they send collaterals to the striatum, thalamus, red nucleus, cerebellum, and reticular formations. Accompanying the corticospinal tracts in the brainstem are the corticobulbar tracts, which are distributed to motor nuclei of the cranial nerves ipsilaterally and contralaterally (Fig. 3-2). It has been possible to trace the direct projection of axons of cortical neurons to the trigeminal, facial, ambiguus, and hypoglossal nuclei (Iwatsubo et al). No axons were seen to terminate directly in the oculomotor, trochlear, abducens, or vagal nuclei. Insofar as the corticobulbar and corticospinal fibers have a similar origin and the motor nuclei of the brainstem are the homologues of the motor neurons of the spinal cord, the term upper motor neurons may suitably be applied to both these systems of fibers.
The corticospinal tracts decussate at the lower end of the medulla, although some of their fibers may cross above this level. The fibers destined for the upper limb neurons cross first (more rostrally). The proportion of crossed and uncrossed fibers varies to some extent from one person to another. About 75 to 80 percent of the fibers cross and the remaining fibers descend ipsilaterally, mostly in the uncrossed ventral corticospinal tract. In exceptional cases, these tracts cross completely; equally rarely, they remain uncrossed. These variations are probably of functional significance in determining the amount of neurologic deficit that results from a unilateral lesion such as capsular infarction. A few well-studied cases are found, such as the one described by Terakawa and colleagues, of acute stroke of the cerebral hemisphere causing hemiplegia on the same side. Also, Yakovlev found 3 instances of completely uncrossed pyramids among 130 autopsies of mentally retarded neonates but considering the maldevelopment of these brains, the finding may not be surprising.
The corticospinal tract is phylogenetically relatively new, being found only in mammals, which probably accounts for its variability between individuals as compared to the older vestibulospinal, rubrospinal and reticulospinalparapyramidal systems, which are invariant among persons. Uncrossed fibers in the corticospinal tract account for mirror movements that are seen during efforts at fine motor tasks, particularly in children, and also in some disorders of the nervous system such as the Klippel-Feil syndrome and the Kallmann syndrome. For a more complete discussion of the crossing of the various tracts of the nervous system, the reader is referred to the review by Vulliemoz, Raineteau, and Jabaudon.
Beyond their decussation, the corticospinal pathways descend as well-defined bundles in the anterior and posterolateral columns of white matter (funiculi) of the spinal cord (Fig. 3-2). The course of the noncorticospinal motor pathways (vestibulospinal, reticulospinal, and descending propriospinal) have been traced in humans by Nathan and his colleagues. The lateral vestibulospinal tract lies at the periphery of the cord, where it occupies the most anterolateral portion of the anterior funiculus. The medial vestibulospinal fibers mingle with those of the medial longitudinal fasciculus. Reticulospinal fibers are less compact; they descend bilaterally, and most of them come to lie just anterior to the lateral corticospinal tract. The descending propriospinal pathway consists of a series of short fibers (one or two segments long) lying next to the gray matter.
The somatotopic organization of the corticospinal system is of importance in clinical work, especially in relation to certain stroke syndromes. As the descending axons subserving limb and facial movements emerge from the cortical motor strip, they maintain the anatomic organization of the overlying cortex; therefore a discrete cortical–subcortical lesion will result in a restricted weakness of the hand and arm or the foot and leg. More caudally, the descending motor fibers converge and are collected in the posterior limb of the internal capsule, so that even a small lesion there will cause a "pure motor hemiplegia," in which the face, arm, hand, leg, and foot are affected to more or less the same degree (see Lacunar syndromes in Chap. 34). The axons subserving facial movement are situated rostrally in the posterior limb of the capsule, those for hand and arm in the central portion and those for the foot and leg, caudally (as detailed by Brodal).
This topographic distribution is maintained in the cerebral peduncle, where the corticospinal fibers occupy approximately the middle of the peduncle, the fibers destined to innervate the facial nuclei lying most medially. More caudally, in the basis pontis (base, or ventral part of the pons), the descending motor tracts separate into bundles that are interspersed with masses of pontocerebellar neurons and their cerebellipetal fibers. A degree of somatotopic organization can be recognized here as well, exemplified by selective weakness of the face and hand with dysarthria, or of the leg, which may occur with pontine lacunar infarctions. Anatomic studies in nonhuman primates indicate that arm–leg distribution of fibers in the rostral pons is much the same as in the cerebral peduncle; in the caudal pons, this distinction is less-well defined. In humans, a lack of systematic anatomic study leaves the precise somatotopic organization of corticospinal fibers in the pons less certain. Restricted pontine lesions may cause a pure motor hemiplegia that is indistinguishable from the syndrome of the internal capsule. However, a study conducted by Marx and colleagues using sophisticated MRI mapping techniques of patients with hemiplegia from brainstem lesions suggests that the usual somatotopic organization breaks down in the base of the pons, and there is a concentration of fibers innervating proximal muscles lying more dorsally and those exciting distal parts of the limbs, more ventrally.
Another point of uncertainty has been the existence and course of fibers that descend through the lower pons and upper medulla and then ascend again to innervate the facial motor nucleus on the opposite side. Such a connection must exist to explain occasional instances of facial palsy from brainstem lesions caudal to the midpons. A discussion of the various hypothesized sites of this pathway, including a recurrent tract (Pick bundle), can be found in the report by Terao and colleagues. They conclude from imaging studies that corticobulbar fibers destined for the facial nucleus descend in the ventromedial pons to the level of the upper medulla, where they decussate and then ascend again; but there is considerable variation between individuals in this configuration.
The descending pontine bundles, now devoid of their corticopontine fibers, reunite to form the medullary pyramid. The brachial–crural pattern may persist in the pyramids and is certainly reconstituted in the lateral columns of the spinal cord (Fig. 8-3), but it should be emphasized that the topographic separation of motor fibers at cervical, thoracic, lumbar, and sacral levels is not as discrete as usually shown in schematic diagrams of the spinal cord.
The corticospinal tracts and other upper motor neurons terminate mainly in relation to nerve cells in the intermediate zone of spinal gray matter (internuncial neurons), from which motor impulses are then transmitted to the anterior horn cells. Only 10 to 20 percent of corticospinal fibers (presumably the thick, rapidly conducting axons derived from Betz cells) establish direct synaptic connections with the large motor neurons of the anterior horns.
Motor, Premotor, and Supplementary Motor Cortices and Cerebral Control of Movement
The motor area of the cerebral cortex is defined physiologically as the region of electrically excitable cortex from which isolated movements can be evoked by stimuli of minimal intensity. The muscle groups of the contralateral face, arm, trunk, and leg are represented in the primary motor cortex (area 4 in Fig. 3-3), those of the face being in the most inferior part of the precentralgyrus on the lateral surface of the cerebral hemisphere and those of the leg in the paracentral lobule on the medial surface of the cerebral hemisphere. The parts of the body capable of the most delicate movements have, in general, the largest cortical representation, as displayed in the motor homunculus ("little man," a term first suggested by Wilder Penfield) shown in Fig. 3-4.
The representation of body parts in the motor cortex, commonly called the motor homunculus. The large area of cortex devoted to motor control of the hand, lips, and face is evident. B in the smaller diagram represents the motor cortex; A is the sensory cortex.
Area 6, the premotor area, is also electrically excitable but requires more intense stimuli than area 4 to evoke movements. Stimulation of its caudal aspect (area 6a) produces responses that are similar to those elicited from area 4. These responses are probably produced by transmission of impulses from all of area 6a to area 4 (as they cannot be obtained after ablation of area 4). Stimulation of the rostral premotor area (area 6a) elicits more general movement patterns, predominantly of proximal limb musculature. The latter movements are effected via pathways other than those derived from area 4 (hence, "parapyramidal"). Very strong stimuli elicit movements from a wide area of premotor frontal and parietal cortex, and the same movements may be obtained from several widely separated points. From this it may be assumed, as Ash and Georgopoulus point out, that the premotor cortex includes several anatomically distinct subregions with different afferent and efferent connections. In general, it may be said that the motor–premotor cortex is capable of synthesizing agonist actions into an almost infinite variety of finely graded, highly differentiated patterns. These are directed by visual (area 7) and tactile (area 5) sensory information and supported by appropriate postural mechanisms.
The supplementary motor area is the most anterior portion of area 6 on the medial surface of the cerebral hemisphere (area 6a in Fig. 3-3B). Stimulation of this area may induce relatively gross ipsilateral or contralateral movements, bilateral tonic contractions of the limbs, contraversive movements of the head and eyes with tonic contraction of the contralateral arm, and sometimes inhibition of voluntary motor activity and vocal arrest.
Precisely how the motor cortex controls movements is still a controversial matter. The traditional view, based on the interpretations of Hughlings Jackson and of Sherrington, has been that the motor cortex is organized not in terms of individual muscles but of movements, i.e., the coordinated contraction of groups of muscles. Jackson visualized a widely overlapping representation of muscle groups in the cerebral cortex on the basis of his observation that a patient could recover the use of a limb following destruction of the limb area as defined by cortical stimulation. This view was supported by Sherrington's observations that stimulation of the cortical surface activated not solitary muscles but a combination of muscles, and always in a reciprocal fashion—i.e., in a manner that maintained the expected relationship between agonists and antagonists. He also noted the inconstancy of stimulatory effects; the stimulation of a given cortical point that initiated flexion of a part on one occasion might initiate extension on another.
These interpretations must be viewed with circumspection, as must all observations based on the electrical stimulation of the surface of the cortex. It has been shown that to stimulate motor cells from the surface, the electric current has to penetrate the cortex to layer V, where these neurons are located, inevitably activating a large number of other cortical neurons. The elegant experiments of Asanuma and of Evarts and his colleagues, who stimulated the depths of the cortex with microelectrodes, demonstrated the existence of discrete zones of efferent neurons that control the contraction of individual muscles; moreover, the continued stimulation of a given efferent zone often facilitated rather than inhibited the contraction of the antagonists. These investigators have also shown that cells in the efferent zone receive afferent impulses from the particular muscle to which the efferent neurons project. When the effects of many stimulations at various depths were correlated with the exact sites of each penetration, cells that projected to a particular pool of spinal motor neurons were found to be arranged in radially aligned columns approximately 1 mm in diameter.
The columnar arrangement of cells in the sensorimotor cortex had been appreciated for many years; the wealth of radial interconnections between the cells in these columns led Lorente de Nó to suggest that these "vertical chains" of cells were the elementary functional units of the cortex. This notion received strong support from Mountcastle's observations that all the neurons in a column receive impulses of the same sensory modality, from the same part of the body. It is still not entirely clear whether the columns contribute to a movement as units or whether individual cells within many columns are selectively activated. Both Henneman and Asanuma summarized the evidence for these disparate views.
Evarts and his colleagues also elucidated the role of cortical motor neurons in sensory evoked or planned movement. Using single-cell recording techniques, they showed that pyramidal cells fire about 60 ms prior to the onset of a movement, in a sequence determined by the required pattern and force of the movement. But other, more complex properties of the pyramidal cells were also noted. Some of them received a somatosensory input transcortically from the parietal lobe (areas 3, 1, and 2), which could be turned on or off or gated according to whether the movement was to be controlled, i.e., guided, by sensory input. Many neurons of the supplementary and premotor cortices were activated before a planned movement. Thus pyramidal (area 4) motor neurons were prepared for the oncoming activation by impulses from the parietal, prefrontal, premotor, and auditory and visual areas of the cortex. This preparatory "set signal" could occur in the absence of any activity in the spinal cord and muscles. The source of the activation signal was found to be mainly in the supplementary motor cortex, which appears to be under the direct influence of the "readiness stimuli" (Bereitschaft potential) reaching it from the prefrontal areas for planned movements and from the posterior parietal cortex for motor activities initiated by sensory perceptions. There are also fibers that reach the motor area from the limbic system, presumably subserving motivation and attention. Roland has used functional cerebral blood flow measurements to follow these neural events.
Thus the prefrontal cortex, supplementary motor cortex, premotor cortex, and motor cortex are all responsive to afferent stimuli and are involved prior to, and in coordinated fashion with, a complex movement. As remarked later on, the striatopallidum and cerebellum, which project to these cortical areas, are also activated prior to or concurrently with the discharge of corticospinal neurons (see Thach and Montgomery for a critical review of the physiologic data).
Termination of the Corticospinal and Other Descending Motor Tracts
This has been studied in the monkey by interrupting the descending motor pathways in the medulla and more rostral parts of the brainstem and tracing the distribution of the degenerating elements in the spinal gray matter. On the basis of such experiments and other physiologic data, Lawrence and Kuypers proposed that the functional organization of the descending cortical and subcortical pathways is determined more by their patterns of termination and the motor capacities of the internuncial neurons upon which they terminate than by the location of their cells of origin. Three groups of motor fibers were distinguished according to their differential terminal distribution: (1) The corticospinal and corticobulbar tracts, which project to all levels of the spinal cord and brainstem, terminating diffusely throughout the nucleus proprius of the dorsal horn and the intermediate zone. A portion of these connect directly with the large motor neurons that innervate the muscles of the fingers, face, and tongue; this system provides the capacity for a high degree of fractionation of movements, as exemplified by independent finger movements.
As alluded to above, a large fraction of the fibers in the corticospinal originate from the sensory cortex and appear to function in the modulation of movement by afferent neurons. (2) A ventromedial pathway, which arises in the tectum (tectospinal tract), vestibular nuclei (vestibulospinal tract), and pontine and medullary reticular cells (reticulospinal tract) and terminates principally on the internuncial cells of the ventromedial part of the spinal gray matter. This system is mainly concerned with axial movements—the maintenance of posture, integrated movements of body and limbs, and total limb movements. (3) A lateral pathway, which is derived mainly from the magnocellular part of the red nucleus and terminates in the dorsal and lateral parts of the internuncial zone. This pathway adds to the capacity for independent use of the extremities, especially of the hands.
Reference has already been made to the corticomesencephalic, corticopontine, and corticomedullary fiber systems that project onto the reticulospinal, vestibulospinal, rubrospinal, and tectospinal nuclei. These control stability of the head (via labyrinthine reflexes) and of the neck and body in relation to the head (tonic neck reflexes) as well as postures of the body in relation to limb movements. Lesions in these systems are less well understood than those of the corticospinal system. They cause no paralysis of muscles but result in the liberation of unusual postures (e.g., hemiplegic dystonia), heightened tonic neck and labyrinthine reflexes, and decerebrate rigidity. In a strict sense these are all "extrapyramidal," as discussed in the next two chapters.
Paralysis Caused by Lesions of the Upper Motor Neurons
The corticospinal pathway may be interrupted by a lesion at any point along its course—at the level of the cerebral cortex, subcortical white matter, internal capsule, brainstem, or spinal cord. Usually, when hemiplegia is severe and permanent as a consequence of disease, much more than the long, direct corticospinal pathway is involved. In the cerebral white matter (corona radiata) and internal capsule, the corticospinal fibers are intermingled with corticostriate, corticothalamic, corticorubral, corticopontine, cortico-olivary, and corticoreticular fibers. It is noteworthy that thalamocortical fibers, which are a vital link in an ascending fiber system from the basal ganglia and cerebellum, also pass through the internal capsule and cerebral white matter. Thus lesions in these parts can simultaneously affect both corticospinal and extrapyramidal systems. Attribution of a capsular hemiplegia solely to a lesion of the corticospinal or pyramidal pathway is therefore not entirely correct. The term upper motor neuron (supranuclear) paralysis, which recognizes the involvement of several descending fiber systems that influence and modify the lower motor neuron, is more appropriate.
In primates, lesions limited to area 4 of Brodmann, the motor cortex, cause mainly hypotonia and weakness of the distal limb muscles. Lesions of the premotor cortex (area 6) result in weakness, spasticity, and increased stretch reflexes (Fulton). Lesions of the supplementary motor cortex lead to involuntary grasping. Resection of cortical areas 4 and 6 and subcortical white matter in monkeys causes complete and permanent paralysis and spasticity (Laplane et al). These clinical effects have not been as clearly defined in humans.
The one place where corticospinal fibers are entirely isolated is the pyramidal tract in the medulla. In humans, there are a few documented cases of a lesion more or less confined to this location. The result of such lesions has been an initial flaccid hemiplegia (with sparing of the face), from which there is considerable recovery. Similarly in monkeys—as was shown by Tower in 1940 and subsequently by Lawrence and Kuypers and by Gilman and Marco—interruption of both pyramidal tracts results in a hypotonic paralysis; ultimately, these animals regain a wide range of movements, although slowness of all movements and loss of individual finger movements remain as permanent deficits. Also, the cerebral peduncle had in the past been sectioned in patients in an effort to abolish involuntary movements (Bucy et al). In some of these patients, a slight degree of weakness or only a Babinski sign was produced but no spasticity developed. These observations indicate that a pure pyramidal tract lesion does not result in spasticity. Furthermore, to reiterate a previous comment, control over a wide range of voluntary movements depends at least in part on nonpyramidal motor pathways. Animal experiments suggest that the corticoreticulospinal pathways are particularly important in this respect, because their fibers are arranged somatotopically and influence stretch reflexes. Further studies of human disease, possibly using diffusion tensor imaging techniques, are necessary to settle problems related to volitional movement and spasticity.
The distribution of the paralysis caused by upper motor neuron (supranuclear) lesions varies with the locale of the lesion, but certain features are characteristic of all of them. A group of muscles is always involved, never individual muscles, and if any movement is possible, the proper relationships between agonists, antagonists, synergists, and fixators are preserved. On careful inspection, the paralysis never involves all the muscles on one side of the body, even in the severest forms of hemiplegia. Movements that are invariably bilateral—such as those of the eyes, jaw, pharynx, upper face, larynx, neck, thorax, diaphragm, and abdomen—are affected little or not at all. This occurs because these muscles are bilaterally innervated; i.e., stimulation of either the right or left motor cortex results in contraction of these muscles on both sides of the body. Upper motor neuron paralysis is rarely complete for any long period of time; in this respect it differs from the absolute paralysis that results from destruction of anterior horn cells or interruption of their axons.
Upper motor neuron lesions are characterized further by certain peculiarities of retained movement. There is decreased voluntary drive on spinal motor neurons (fewer motor units are recruitable and their firing rates are slower), resulting in a slowness of movement. There is also an increased degree of co-contraction of antagonistic muscles, reflected in a decreased rate of rapid alternating movements. These abnormalities probably account for the greater sense of effort and the manifest fatigability in effecting voluntary movement of the weakened muscles. Another phenomenon is the activation of paralyzed muscles as parts of certain automatisms (synkinesias). For example, the paralyzed arm may move suddenly during yawning and stretching. Attempts by the patient to move the hemiplegic limbs may also result in a variety of associated movements. Thus, flexion of the arm may result in involuntary pronation and flexion of the leg or in dorsiflexion and eversion of the foot. Also, volitional movements of the paretic limb often evoke imitative (mirror) movements in the normal one or vice versa. Mirror movements are also a feature of Parkinson disease and of lesions in the upper cervical spinal cord. In some patients, as they recover from hemiplegia, a variety of movement abnormalities emerge, such as tremor, athetosis, and chorea on the affected side. These are expressions of damage to basal ganglionic and thalamic structures and are discussed in Chap. 4.
If the upper motor neurons are interrupted above the level of the facial nucleus in the pons, hand and arm muscles are affected most severely and the leg muscles to a lesser extent; of the cranial musculature, only muscles of the tongue and lower part of the face are involved to any significant degree. Because Broadbent was the first to call attention to this distribution of facial paralysis that relatively spares the forehead muscles, it is referred to as "Broadbent's law." The precise course taken by fibers that innervate the facial nucleus is still somewhat uncertain; however, the majority crosses in the mid-pons to innervate the contralateral facial nerve nucleus. Some fibers may descend to the upper medulla and then ascend recurrently to the pons, (Pick's bundle accounting for the mild facial weakness that is seen with lesions of the lower pons and upper medulla.
At lower levels, such as the cervical cord, complete, acute, and bilateral lesions of the upper motor neurons not only cause a paralysis of voluntary movement but also temporarily abolish the spinal reflexes of segments below the lesion. This is the condition referred to earlier as spinal shock, a state of acute flaccid paralysis that is replaced later by spasticity. A comparable state of areflexia and hypotonia may occur with acute cerebral lesions but is less sharply defined than is the spinal state. With some acute cerebral lesions, spasticity and paralysis develop together; in others, especially with parietal lesions, the limbs remain flaccid but reflexes are retained.
Spasticity, Hyperreflexia, and the Babinski Sign
The identifying characteristics of paralysis from an upper motor neuron lesion are a predilection for involvement of certain muscle groups, a specific pattern of response of muscles to passive stretch (where resistance increases linearly in relation to velocity of stretch, and a manifest exaggeration of tendon reflexes. The antigravity muscles—the flexors of the arms and the extensors of the legs—are predominantly affected. The arm tends to assume a flexed and pronated position and the leg an extended and adducted one, indicating that certain spinal neurons are reflexly more active than others. At rest, with the muscles shortened to midposition, they are flaccid to palpation and electromyographically silent. If the arm is extended or the leg flexed very slowly, there may be little or no change in muscle tone. By contrast, if the muscles are briskly stretched, the limb moves freely for a very short distance (free interval), beyond which there is an abrupt catch and then a rapidly increasing muscular resistance up to a point; then, as passive extension of the arm or flexion of the leg continues, the resistance melts away. This velocity dependent tone constitutes the "clasp-knife" phenomenon of spasticity. With the limb in the extended or flexed position, a new passive movement may not encounter the same sequence; this entire pattern of response constitutes the lengthening and shortening reaction. Thus, the essential feature of spasticity is a velocity-dependent increase in the resistance of muscles to a passive stretch stimulus.
Although a clasp-knife relaxation following peak resistance is highly characteristic of cerebral hemiplegia, it is by no means found consistently. At times, a form of velocity-independent hypertonia is found that is termed rigidity and is more characteristic of basal ganglia lesions as discussed in Chap. 4.
Clinicians have known for some time that there is not a constant relationship between spasticity and weakness. Severe weakness may be associated with only the mildest signs of spasticity; in contrast, the most extreme degrees of spasticity, observed in certain patients with cervical spinal cord disease, may seem disproportionate to the extent of weakness, signifying that these two states depend on separate mechanisms. Indeed, the selective blocking of small gamma neurons abolishes spasticity as well as hyperactive segmental tendon reflexes but to leave power unchanged.
The heightened stretch reflexes (tendon jerks) of the spastic state may be a "release" phenomenon—the result of interruption of descending inhibitory pathways, but this appears to be only a partial explanation. Animal experiments have demonstrated that this aspect of the spastic state is also mediated through spindle efferents (increased tonic activity of gamma motor neurons) and, centrally, through reticulospinal and vestibulospinal pathways that act on alpha motor neurons. The clasp-knife phenomenon appears to derive at least partly from a lesion (or presumably a change in central control) of a specific portion of the reticulospinal system.
P. Brown, in a discussion of the pathophysiology of spasticity, emphasized the importance of two systems of fibers: (1) the dorsal reticulospinal tract, which has inhibitory effects on stretch reflexes; and (2) the medial reticulospinal and vestibulospinal tracts, which together facilitate extensor tone. He postulated that in cerebral and capsular lesions, cortical inhibition is reduced, resulting in spastic hemiplegia. In spinal cord lesions that involve the corticospinal tract, the dorsal reticulospinal tract is usually involved as well. If the latter tract is spared, only paresis, loss of support reflexes, and possibly release of flexor reflexes (Babinski phenomenon) occur. Pantano and colleagues suggested that primary involvement of the lenticular nucleus of the basal ganglia and thalamus is the feature that determines the persistence of flaccidity after stroke, but the anatomic and physiologic evidence for this view is insecure.
The most sensitive indications of an upper motor neuron lesion are the signs described by Babinski in 1896 (the great toe sign) and 1903 (the toe abduction, or fan sign). In modern parlance, the toe and fan signs have generally been conflated and termed the Babinski sign. Numerous monographs and articles have been written about the sign: a quite comprehensive one, by van Gijn, and an elegant but more arcane one by Fulton and Keller.
As Babinski himself indicated, a movement resembling the Babinski sign is present in normal infants (see Phiilipon and Poirer), but it disappears and its persistence or emergence in late infancy and childhood or later in life is an invariable indicator of a lesion at some level of the corticospinal tract. In its essential form, the sign consists of extension of the large toe and extension and fanning of the other toes during and immediately after stroking the lateral plantar surface of the foot. The stimulus is applied along the dorsum of the foot from the lateral heel and sweeping upward and across the ball of the foot. The stimulus must be firm but not necessarily painful. Several dozen surrogate responses (with numerous eponyms) have been described over the years, most utilizing alternative sites and types of stimulation, but all have the same significance as the Babinski response.
Clinical and electrophysiologic observations indicate that the extension movement of the toe is a component of a larger synergistic flexion or shortening reflex of the leg—i.e., toe extension when viewed from a physiologic perspective is a flexor protective (nocifensive, or defensive) response. The most characteristic of these is the "triple flexion response", in which the hip, thigh and ankle flex (dorsiflex) slowly, following an appropriate stimulus. These spinal flexion reflexes, of which the Babinski sign is the most characteristic, are common accompaniments to—but not essential components of—spasticity. They are present because of disinhibition or release of motor programs of spinal origin. Important characteristics of these responses are their capacity to be induced by weak superficial stimuli (such as a series of pinpricks) and their tendency to persist for a few moments after the stimulation ceases. With incomplete suprasegmental lesions, the response may be fractionated; for example, the hip and knee may flex but the foot may not dorsiflex, or vice versa.
The hyperreflexic state that characterizes spasticity may take the form of clonus, a series of rhythmic involuntary muscular contractions occurring at a frequency of 5 to 7 Hz in response to an abruptly applied and sustained stretch stimulus. It is usually designated in terms of the part of the limb to which the stimulus is applied (e.g., patella, ankle). The frequency is constant within 1 Hz and is not appreciably modified by altering peripheral or central nervous system activities. Clonus requires an appropriate degree of muscle relaxation, integrity of the spinal stretch reflex mechanisms, sustained hyperexcitability of alpha and gamma motor neurons (suprasegmental effects), and synchronization of the contraction–relaxation cycle of muscle spindles.
The cutaneomuscular abdominal and cremasteric reflexes ("cutaneous, or superficial reflexes") are elicited by rapid, gentle stroking of the skin overlying these muscles, and are usually abolished when the upper motor neuron is damaged. These were referred to as reflexes before the end of the nineteenth century, which leads to some confusion in interpreting the older clinical literature.
Spread, or radiation of reflexes, is regularly associated with spasticity, although the latter phenomenon may be observed to a slight degree in normal persons with brisk tendon reflexes. Tapping of the radial periosteum, for example, may elicit a reflex contraction not only of the brachioradialis but also of the biceps, triceps, or finger flexors. This spread of reflex activity is probably not the result of radiation of impulses in the spinal cord, but a result of the propagation of a vibration wave from bone to muscle, stimulating the excitable muscle spindles in its path (Lance). Other manifestations of the hyperreflexic state, are the Hoffmann sign and the crossed adductor reflex of the thigh muscles. Also, reflexes may be "inverted," as in the case of a lesion of the fifth or sixth cervical segment; here the biceps and brachioradialis reflexes are abolished and only the triceps and finger flexors, whose reflex arcs are intact, respond to a tap over the distal radius.
With bilateral cerebral lesions, exaggerated stretch reflexes may be elicited in cranial as well as limb and trunk muscles because of interruption of the corticobulbar pathways. These are seen as easily triggered masseter contractions in response to a brisk downward tap on the chin ("jaw jerk") and brisk contractions of the orbicularis oris muscles in response to tapping the philtrum or corners of the mouth. In advanced cases, weakness or paralysis or slowness of voluntary movements of the face, tongue, larynx, and pharynx are added (bulbar spasticity or "pseudobulbar" palsy; see also Chap. 25).
The many investigations of the biochemical changes that underlie spasticity and the mechanisms of action of antispasticity drugs have been reviewed by Davidoff. Because glutamic acid is the neurotransmitter of the corticospinal tracts, one would expect its action on inhibitory interneurons to be lost. As mentioned earlier, GABA and glycine are the major inhibitory transmitters in the spinal cord; GABA functions as a presynaptic inhibitor, suppressing sensory signals from muscle and cutaneous receptors. Baclofen, a derivative of GABA, as well as diazepam and progabide, are thought to act by reducing the release of excitatory transmitters from the presynaptic terminals of primary afferent terminals. Actually, none of these agents is entirely satisfactory in the treatment of spasticity when administered orally; the administration of baclofen intrathecally at times has a more beneficial effect. Glycine is the transmitter released by inhibitory interneurons and is measurably reduced in quantity, uptake, and turnover in the spastic animal. There is some evidence that the oral administration of glycine reduces experimentally induced spasticity, but its value in patients is uncertain. Interruption of descending noradrenergic, dopaminergic, and serotonergic fibers is undoubtedly involved in the genesis of spasticity, although the exact mode of action of these neurotransmitters on the various components of spinal reflex arcs remains to be defined.
Table 3–1 summarizes the main attributes of upper motor neuron lesions and contrasts them with those of the lower motor neuron discussed above.
Table 3–1 Differences between Upper and Lower Motor Neuron Paralysis ||Download (.pdf)
Table 3–1 Differences between Upper and Lower Motor Neuron Paralysis
UPPER MOTOR NEURON OR SUPRANUCLEAR PARALYSIS
LOWER MOTOR NEURON OR NUCLEAR-INFRANUCLEAR PARALYSIS
Muscles affected in groups; never individual muscles
Individual muscles may be affected
Atrophy slight and the result of disuse
Atrophy pronounced; up to 70% of total bulk
Spasticity with hyperactivity of the tendon reflexes and extensor plantar reflex (Babinski sign)
Flaccidity and hypotonia of affected muscles with loss of tendon reflexes
Plantar reflex, if present, is of normal flexor type
Normal nerve conduction studies; no denervation potentials in EMG
Fasciculations may be present
Abnormal nerve conduction studies; denervation potentials (fibrillations, fasciculations, positive sharp waves) in EMG
Motor Disturbances Caused by Lesions of the Parietal Lobe
As indicated earlier in this section, a significant portion of the pyramidal tract originates in neurons of the parietal cortex. Also, the parietal lobes are important sources of visual and tactile information necessary for the control of movement. Pause and colleagues have described the motor disturbances caused by lesions of the parietal cortex. The patient is unable to maintain stable postures of the outstretched hand when his eyes are closed and cannot exert a steady contraction. Exploratory movements and manipulation of small objects are impaired, and the speed of tapping is diminished. Posterior parietal lesions (involving areas 5 and 7 in Fig. 3-3) are more detrimental in this respect than anterior ones (areas 1, 3, and 5), but both regions are affected in patients with the most severe deficits.
1Numbered areas in this and subsequent chapters refer to Brodmann areas of the cerebral cortex that are discussed in Chap. 23. "Layers" refer to the six neuronal layers of the cerebral cortex, also shown in detail in Chap. 23, on Cerebral Localization.