CHAPTER 10 Small Animal Spinal Cord Disease
The objective of this chapter is first to review the method and interpretation of the neurologic examination as it relates to localizing lesions in the spinal cord in small animals, and second, to present case studies that illustrate examples of these in the four regions of the spinal cord.
Careful examination of all of these is necessary to determine whether a lesion is confined to the spinal cord and at what level. Remember that, as described in Chapter 5, spinal reflexes require only the specific peripheral nerves and the spinal cord segments with which they connect, whereas postural reactions depend on the same components as the spinal reflexes plus the cranial projecting general proprioceptive (GP) pathways in the spinal cord white matter to the brainstem, cerebellum, and frontoparietal cerebrum and the caudal projecting upper motor neuron (UMN) pathways that return from the cerebrum and brainstem and comprise tracts in the white matter of the spinal cord that terminate in the cervical and lumbosacral intumescences. These postural reactions test the integrity of nearly the entire peripheral and central nervous systems. By themselves, postural reactions are relatively nonlocalizing for a lesion.
The clinical signs of UMN and GP dysfunction are described in Chapters 8 and 9, and it is emphasized that most gait abnormalities involving these systems reflect a combination of UMN (spastic paresis) and GP (ataxia) clinical signs because of their close anatomic relationship. It also is strongly urged that differentiation between the clinical signs caused by disruption of these two systems is of no practical value. With spinal cord lesions that affect the UMN and GP systems between spinal cord segments T3 and L3, there is a tendency to describe the pelvic limb gait as just ataxic or occasionally just paretic. In reality, the gait abnormality usually reflects a dysfunction of both systems and should be referred to as pelvic limb ataxia and paresis or paraparesis and pelvic limb ataxia. The same rule holds for cervical spinal cord lesions that occur between spinal cord segments C1 and C5: tetraparesis (quadriparesis) and ataxia of all four limbs. The terms paraplegia and tetraplegia should be used only when there is absolutely no voluntary movement in the pelvic limbs or in all four limbs, respectively. To observe ataxia, there has to be voluntary movement. Therefore, the term ataxia is inappropriate for a paraplegic or tetraplegic animal. Patients that are recumbent in the pelvic limbs (T3-L3 lesion) or in all four limbs (C1-C5 lesion) and show no voluntary movement when picked up and moved along the ground surface while suspended but exhibit some voluntary limb movement when recumbent should be described as nonambulatory paraparetic or tetraparetic. The same terminology applies to lesions that involve the lumbosacral or cervical intumescence, except that the quality of the paresis reflects a lower motor neuron (LMN) dysfunction.
The gait should be examined in a place where the patient can move freely, leashed or unleashed, and where the ground surface is not slippery. The floor of many examining rooms is too small and too slippery for an adequate evaluation of the gait. In some patients with vertebral column injury and spinal cord contusion that are ataxic and paretic, moving the patient on a slippery surface may cause it to fall, and further injury may result. The availability of a corridor and an indoor-outdoor carpet is very helpful for this examination.
I (Eric Glass) use a covered outdoor area to evaluate neurologic patients. It has a specialized surface that is used in many playgrounds. This surface is soft and provides excellent traction for paretic and ataxic patients. The material, Vitriturf (Hanover Specialties, Hauppauge, NY), is commercially available and easily installed. It is important to evaluate the gait while you lead the patient and while an assistant leads the patient, in a straight line as well as in circles in each direction.
A patient with lesions that affect the pontomedullary or spinal cord UMN and GP systems and is still able to walk unassisted exhibits a delay in the onset of protraction, which is the swing phase of the gait; a stiff quality of movement; and often a longer stride, due to a delay in the termination of protraction. In the thoracic limbs, this causes a floating, overreaching action, with the limb in extension. It is important to recognize the extension of the thoracic limb joints and the prolongation of the protraction, which causes the paw to be placed on the ground farther cranially than normal. This is in contrast to the thoracic limb gait in a cerebellar disorder, where the protraction is abrupt and all the limb joints flex, causing the limb to move more dorsally toward the neck. As protraction is completed, the limb is commonly misdirected. This difference in the thoracic limb gait between a UMN-GP system dysfunction and a cerebellar dysfunction is difficult to describe but should be obvious when you study the videos of these disorders. (See Chapter 13 for a description of the gait of a cerebellar disorder.) In UMN-GP system disorders, on protraction, the affected limb may abduct excessively, especially on turns, when it is often referred to as circumduction. The limb may also adduct excessively before it supports weight. This may appear as a crossing of the limbs as they are advanced. Often, the dorsal surface of the paw will scuff the ground on protraction or the patient will support its weight on the dorsal surface of the paw. Occasionally a pelvic limb is flexed excessively on protraction, creating a hypermetria in the gait. The trunk may also appear to be unstable and to sway as the patient walks, especially on turns. Remember that these clinical signs reflect dysfunction in both the UMN and the GP systems. There is ample opportunity in the disease section of this chapter to visualize on videos what is described here.
Many clinicians use a grading system to assess the degree of pelvic limb function as an aid in determining the prognosis and to evaluate response to therapy. This was originally designed for dogs with thoracolumbar spinal cord injury resulting from intervertebral disk extrusions. This is a common disorder in dogs in which numerous surgical procedures have been performed over many decades. However, it is applicable to any spinal cord lesion at any level. Grade 0 refers to a patient with no voluntary movement in the pelvic limbs and therefore is paraplegic. Grade 5 refers to a patient with normal pelvic limb function. Patients with functional grades 1 and 2 are unable to stand in the pelvic limbs without assistance. When you hold the patient up by grasping the base of the tail, and only very slight movements of the pelvic limbs occur, this patient would receive a grade of 1 for its function. If voluntary movements readily occur but are delayed, awkward, and poorly placed, the degree of function would be grade 2. A patient with grade 3 function is able to stand up in its pelvic limbs without assistance but has great difficulty and is able to walk but with significant paresis and ataxia. A patient with grade 4 function readily stands up unassisted and exhibits only mild paresis and ataxia in its gait. These are grades of function, not grades of paresis and ataxia. When we describe a grading system that is used for horses (see Chapter 11), it refers to grades of dysfunction and is used primarily for horses with cervical spinal cord disease.
The degree of functional deficit dictates the need for postural reaction testing. The postural reactions are always absent in all limbs of a tetraplegic patient and in the pelvic limbs of a paraplegic patient. However, in the paraplegic patient it is very important to evaluate carefully the gait and postural reactions in the thoracic limbs so as to avoid missing a multifocal disorder. For example, a young dog may present to you with a progressive pelvic limb dysfunction that has become paraplegic, but you determine it also has hopping deficits in the thoracic limbs. This strongly suggests a multifocal anatomic diagnosis and a presumptive diagnosis of an inflammatory disorder such as canine distemper myelitis.
You were introduced to postural reactions in Chapter 5 as they relate to examining patients with neuromuscular disorders. It is critical to remember that postural reactions essentially require that all components of the peripheral nervous system and the central nervous system (CNS) that affect the limb you are testing be intact.
There is no test for just the UMN or just the GP system. Although a number of these postural reactions are described in the neurologic literature, in our experience the most reliable postural reaction is the patient’s ability to hop laterally on one limb. After that would be the paw replacement test. Hemiwalking is most useful when your patient is too large for you to be comfortable evaluating the hopping responses. Placing of the limbs is useful only in the small-sized patients and when the hopping responses are equivocal.
The following is the sequence of my (Alexander de Lahunta) examination of the patient after evaluating its gait. Be sure to know the name of your patient and talk to it continuously to ensure its cooperation. Make its environment as nonthreatening as possible. In the hospital, I always performed my examinations on a rug or a floor with a rubber surface. Straddle the patient, with both you and the patient facing in the same direction, and palpate all of its neck, trunk, and limb muscles from cranial to caudal to determine whether there is any atrophy. After you palpate each limb, flex and extend it to determine both the muscle tone and the range of motion of the joints. If there is significant joint disease that restricts the range of motion, it contributes to the gait disorder. When you place the limb back in its supporting position, place it on the dorsal aspect of the paw and observe how readily the patient returns the paw to its normal position. This is the paw replacement test. Remember, despite what you may hear from your colleagues or read in the literature, that this is not a test for conscious proprioception. No such test exists. Be aware that some dogs are normally quite slow to replace the paw.
To test thoracic limb hopping, with your left hand, elevate the abdomen just enough to take the weight off the pelvic limbs. Put your left elbow on your left thigh to take the stress off your back. With your right hand, pick up the right thoracic limb and hop the patient on its left thoracic limb to the left. You should not have to move your pelvic limbs. Stand still. When you have hopped the patient to the left as far as it is comfortable, change hands so that your right hand elevates the patient’s abdomen and pelvic limbs and your right elbow is resting on your right thigh. With your left hand, pick up the patient’s left thoracic limb and hop the patient to the right on its right thoracic limb. Keep repeating this back and forth until you are comfortable that the responses are normal or abnormal. As you stand over the patient looking down the lateral aspect of the limb that is being hopped, as soon as the shoulder region moves laterally over the paw, the paw should move. Any delay in this is abnormal. The hopping movements should be smooth and not irregular or excessive. The paw should never drag or land on its dorsal surface. Carefully compare one thoracic limb with the other.
For pelvic limb hopping in patients that are not too large, stand beside the patient’s left side and place your left forearm and hand between its thoracic limbs so that you can elevate the thorax to take the weight off its thoracic limbs. With your right hand, pick up the patient’s left pelvic limb and push the patient to its right side so that it has to hop on the right pelvic limb. Then change sides and repeat this for the left pelvic limb. The responses should be brisk and smooth but will not be quite as fast as in the thoracic limbs. Abnormalities are the same as those described for the thoracic limbs. In large patients, the same observations may be made by just standing beside the patient, picking up the pelvic limb on that side, and pushing the pelvic region away from you so that the patient has to hop on the opposite pelvic limb. This can usually be done with the patient standing still on its thoracic limbs. Repeat this on each side until you are comfortable with the responses observed. As you perform these hopping responses, you will also appreciate the degree of muscle tone that is present in the limbs.
In very large patients, these same hopping responses can be evaluated by hemiwalking the patient. Stand beside the patient and pick up both the thoracic and pelvic limbs on that side. Push the patient away from you and observe the hopping responses in both limbs on the opposite side. Change sides and repeat this on the opposite side, and be sure that you compare the thoracic limbs with each other and the pelvic limbs with each other.
In small patients in which the hopping responses are difficult to interpret, it may be useful to test the placing response. Pick the patient up and bring its thoracic limbs to the edge of a table or chair so that the dorsal surface of the paw contacts the front surface of the object. The normal patient will immediately place its paws on the horizontal surface of the table or chair. Test both limbs at the same time as well as individually. Repeat this for the pelvic limbs. Be sure to test this response while holding the patient from each side. Occasionally, for some unknown reason, a normal patient does not respond with the limbs on the side on which it is being held. It may also be useful to block the patient’s vision during this test by extending its neck so that it cannot see the projecting surface of the object you are using for the test.
In cooperative patients, when you are not sure of the thoracic limb function, it may help to wheelbarrow the patient with its neck held in extension. With your right (or left) forearm, elevate the abdomen of your patient so that the pelvic limbs are no longer supporting weight. With your other hand, hold the neck in extension and move the patient forward. In this posture, vision is compromised and there is a greater need for general proprioceptive function. In cases of mild nervous system lesions, this may cause the animal to scuff the dorsal surface of its paws or to overreach on protraction.
The patient should be placed in lateral recumbency. Usually you will need an assistant to help hold the patient in this position. For cats and toy breeds of dogs, I (AD) find it useful to sit on the floor with my back against the wall and my knees flexed, with the patient placed between my thighs and with its back resting on my thighs. This is useful not only to evaluate muscle tone and spinal reflexes but also to perform the cranial nerve examination. Most patients tolerate this position quite well, and it allows you to control the patient easily.
Flex and extend each limb to further appreciate the degree of muscle tone that you already assessed in the standing position and during the testing of the postural reactions. The only tendon reflex that I (AD) perform is the patellar reflex, which is an assessment of the femoral nerve and spinal cord segments L4, L5, and L6. The detailed anatomy of these tendon reflexes is described in Chapter 5. These tests should be performed before testing the flexor reflexes, in which a noxious stimulus is used. Test the flexor reflexes using a pair of forceps for compressing the skin at the base of the digits. Initially, exert just enough compression to obtain a flexor response or evidence of nociception. These flexor reflexes test various components of the spinal cord segments that comprise the lumbosacral or cervical intumescence and the related peripheral nerves that supply the limb being tested. The detailed anatomy of these reflexes is described in Chapters 5 and 9. Remember that these flexor reflexes test not only the reflex arc but also determine the integrity of the nociceptive pathway to the somesthetic cerebral cortex. The latter is very important in evaluating the level of severe focal spinal cord lesions as well as in making a prognosis for those lesions. Be sure to test all of these reflexes with the patient in both left and right recumbency.
When this testing has been completed and the patient is returned to the standing position, perform the cutaneous trunci reflex. (You may see this reflex referred to as the panniculus reflex, but that is a misnomer because the panniculus adiposus is a layer of fat in the trunk region and is not responsible for this cutaneous reflex; see Fig. 5-9.) Using your forceps, gently probe or squeeze the skin along the dorsal midline, starting at the level of the pelvis, and repeat this stimulus over each vertebra until you observe the cutaneous trunci contracting on both sides. In many normal small animals this may not occur until about the midlumbar region, and in a very few normal animals, it will not occur at all. The forceps pressure stimulates impulses in the dorsal branches of the spinal nerves that supply the area stimulated. Because of the short caudal distribution of these dorsal branches, each spinal nerve supplies the skin for a distance of about two vertebrae caudal to the intervertebral foramen, where the spinal nerve emerges from the vertebral canal. The general somatic afferent (GSA) neurons that are stimulated synapse in the respective dorsal gray column on long interneurons whose axons enter the adjacent fasciculus proprius bilaterally but predominantly on the contralateral side, in our opinion. Here these axons course cranially to the C8 and T1 spinal cord segments to terminate in the ventral gray columns by synapsing on the general somatic efferent (GSE) neurons that innervate the cutaneous trunci via the brachial plexus and the lateral thoracic nerve. Recall that this reflex is lost in injuries that cause avulsion of the roots of the spinal nerves that supply the brachial plexus. In these patients, compression of the skin at any level along the trunk elicits only a cutaneous trunci reflex on the side that is opposite from the affected thoracic limb. This reflex is also helpful in determining the level and prognosis of a severe transverse spinal cord lesion such as an injury resulting from a vertebral column fracture or an intervertebral disk extrusion. Similar to nociception, it requires a bilateral transverse lesion to interfere with the pathway of these long interneurons. This cutaneous reflex is especially helpful in patients that are very stoic and respond poorly to most any noxious stimulus to any part of their body.
The examination of cranial nerves is described in Chapters 6 and 7. When presented with a patient that exhibits clinical signs of spinal cord disease, the cranial nerve examination is important for determining whether the spinal cord signs are part of a multifocal disease process. It is also important to determine whether the spinal cord lesion is at a level where it can interfere with the sympathetic pathway to the head and cause Horner syndrome. That syndrome is observed when the cranial nerves are examined.
Most compressive lesions that affect the cranial cervical spinal cord segments cause more obvious clinical signs in the pelvic limbs than in the thoracic limbs. Explanations for this include the following: lesions that compress from the periphery of the spinal cord affect primarily the superficial tracts, which contain cranially coursing GP pathways from the pelvic limbs (Fig. 10-2); the pelvic limbs are further removed from the center of gravity, which is just caudal to the thoracic limbs; more UMN pathways terminate in the cervical intumescence than in the lumbosacral intumescence. Occasionally a cervical spinal cord lesion causes more obvious clinical signs in the thoracic limbs. This is seen with caudocervical lesions that affect the gray matter in the intumescence, causing LMN signs in the thoracic limbs. With more craniocervical lesions, this occurs when the spinal cord lesion affects the tracts closer to the gray matter and spares the more superficial tracts. These are predominantly UMN pathways to the cervical intumescence. A glial neoplasm that arises in the gray matter and grows peripherally toward the surface of the spinal cord will do this as will a midline intervertebral disk protrusion that compresses the spinal cord dorsally and laterally, making the spinal cord form a tent shape over the compressing mass. This disparity in clinical signs is also seen in many dogs with an atlantoaxial subluxation.
Figure 10-2 Transverse section of cervical spinal cord. The more superficial location of pelvic limb spinocerebellar tracts may explain the more profound pelvic limb ataxia that is often observed with compressive lesions. CCB, Cuneocerebellar tract; FC, fasciculus cuneatus; FG, fasciculus gracilis; SCB, spinocerebellar tracts; UMN, upper motor neuron.
Be aware that lesions that affect the first two or possibly three cervical spinal cord segments may also cause clinical signs of vestibular system dysfunction. This presumably results from the interruption of the spinovestibular tracts that carry GP impulses from the first three cervical spinal nerves, which are important in the orientation of the head to the neck.
There are two additional important concepts to remember when dealing with spinal cord lesions. First, lesions in the intumescences that affect the gray matter as well as the white matter present with the clinical signs of loss of the gray matter; that is, LMN clinical signs. You cannot observe UMN or GP clinical signs unless the LMN is intact. Second, a focal lesion between the C1 and C5 or T3 and L3 spinal cord segments that affects both the white matter and the gray matter causes clinical signs referable only to the white matter lesion; that is, UMN and GP clinical signs. Loss of the gray matter in a segment that innervates only the axial muscles cannot be determined by a physical neurologic examination. Only a careful electromyographic study may reveal the denervation of a segment of axial muscles.
Bladder and occasionally rectal dysfunction commonly accompanies spinal cord lesions. These lesions were described, along with the anatomy involved, in Chapter 7. Lesions in the sacral segments result in LMN incontinence in urine and feces; UMN incontinence is common with severe lesions of the cervical and thoracolumbar spinal cord segments. This incontinence is a primary concern in your treatment of the patient.
These two separate clinical disorders often present at the same time in a patient and appear to be exceptions to the rules that we have established for UMN and LMN disorders. They occur only with peracute, usually transverse, spinal cord lesions between the T3 and L3 spinal cord segments. Fractures of the vertebral column are the most common cause of such lesions. Others include infarction caused by fibrocartilaginous emboli and the myelopathy associated with peracute intervertebral disk extrusions. These patients exhibit persistent severe extension of the thoracic limbs in most postures because of disinhibition of the extensor motor neurons in the cervical intumescence. However, when these patients are held up, the thoracic limb gait is normal except for a mild stiffness. These clinical signs represent the Schiff-Sherrington syndrome.79 The disinhibition is not the result of dysfunction in a UMN pathway, which is why these patients can walk so well with the thoracic limbs when the trunk and pelvic limbs are supported. The disinhibition is the result of a sudden loss of the axons in a long interneuronal pathway that originates from neuronal cell bodies primarily in the gray matter of the L1 to L5 spinal cord segments. These interneurons are referred to as border cells because they are located in the dorsolateral border of the ventral gray column of the lumbar spinal cord segments.89 Their axons course cranially in the fasciculus proprius and terminate by synapsing on thoracic limb extensor LMNs in the cervical intumescence (Fig. 10-3). Their normal function is to inhibit these extensor motor neurons. This extensor release phenomenon is observed only with peracute severe lesions, and it spontaneously resolves in about 10 to 14 days. The presence of the Schiff-Sherrington syndrome indicates a severe lesion and a guarded prognosis but does not indicate that recovery cannot occur.
The severe peracute thoracolumbar spinal cord lesion that is responsible for the Schiff-Sherrington syndrome usually causes paraplegia because of the complete interruption in function of the UMN pathways, and it usually causes analgesia caudal to the lesion because of interruption of the nociceptive pathways. As a rule, with progressive transverse spinal cord lesions, paraplegia occurs before analgesia, suggesting that the nociceptive pathways are the most resistant to spinal cord compression and ischemia. However, in these severe peracute transverse thoracolumbar spinal cord lesions in which the Schiff-Sherrington syndrome is present, there usually is severe pelvic limb hypotonia. If you examine this patient within a few hours of the onset of the paraplegia, there may be absent or very depressed pelvic limb tone and spinal reflexes. These paradoxical LMN-like pelvic limb signs in a patient with a UMN pathway interruption represent what is called spinal shock.88 In primates, spinal shock causes areflexia and atonia for 2 to 3 weeks. In domestic animals, the areflexia is observed for only a few hours after the onset of the lesion, but the hypotonia persists for 10 to 14 days, when it is replaced by normal tone at first and then by hypertonia. The reasons a UMN lesion causes LMN signs are poorly understood. One explanation is that the sudden loss of the UMN synapses, indirectly via interneurons or directly on the dendritic zones or the cell bodies of the alpha motor neurons, causes such a disruption to that LMN cell body that it cannot function for a variable period of time. In primates, more of these UMN pyramidal system synapses are directly on the LMN, which may explain the difference in reaction among species that is observed here. Some studies have found an excessive accumulation of the inhibitory neurotransmitter glycine in the lumbosacral intumescence in these patients.86 The basis for the release of glycine is unknown. It is important to understand this unique combination of clinical signs that results from a focal lesion in the UMN and GP systems and not make an anatomic diagnosis of a multifocal disorder.
In a recumbent patient that has a fracture and should not be manipulated, the basis for these clinical signs can be determined through minimal handling of the patient. With the patient lying on the floor, a table, or a stretcher, provide a rigorous noxious stimulus to a digit of a pelvic limb; no movement occurs or there is just a mild reflex flexion if it is a few hours after the onset of clinical signs. Note the bilateral pelvic limb hypotonia. Provide very minimal compression of a digit of one of the hyperextended thoracic limbs or just a light squeeze of that forepaw with your hand, and note the immediate vigorous voluntary withdrawal of the limb. This tells you that you have a transverse thoracolumbar spinal cord lesion with Schiff-Sherrington syndrome in the thoracic limbs and spinal shock in the pelvic limbs. By providing forceps compression of the skin on the midline of the back as a noxious stimulus, starting in the caudal lumbar region and progressing cranially, a line of analgesia and/or a cutaneous trunci reflex can be found that locates the focal lesion, and imaging studies can be pursued with no further manipulation of the patient. If the patient does not have a fracture, its ability to walk with the thoracic limbs when supported differentiates the Schiff-Sherrington syndrome from a severe cervical spinal cord lesion.
(For a complete description of disorders that affect the lumbosacral spinal cord segments or the peripheral nerves associated with these segments, see the case examples of neuromuscular disease in Chapter 5.)
SMALL ANIMAL THORACOLUMBAR SPINAL CORD DISEASES
Examination: Video 10-1 shows the flaccid paraplegia and the hyperextended thoracic limbs but normal voluntary use of these limbs. Note the atonia and areflexia in the left pelvic limb and the mild hypotonia and hyporeflexia in the right pelvic limb. Note the analgesia of the left pelvic limb except for the proximal medial thigh area. Not shown are the normal tail, anal tone, and perineal reflex.
The thoracic limbs exhibit the Schiff-Sherrington syndrome resulting from the sudden loss of function of the lumbar spinal cord neuronal cell bodies of the long interneurons that normally inhibit thoracic limb extensor motor neurons. However, the pelvic limbs’ clinical signs are not due to spinal shock because they are so asymmetric, and the left pelvic limb areflexia had persisted for 24 hours when this examination was done. The intact nociception from the proximomedial thigh area indicates sparing of the genitofemoral nerve, which arises from the L3 and L4 spinal cord segments. The right pelvic limb paralysis represents a mixture of loss of function of the UMN and GP systems in the white matter and mild loss of GSE LMN function, all of which could occur in these segments.
External injury was excluded by the history. Based on the acute onset of clinical signs and the signalment, FCEM and intervertebral disk extrusion are the two most presumptive diagnoses. The absence of any discomfort is more likely with FCEM, but exceptions are common. It is important to make this distinction in the clinical diagnosis because an extruded intervertebral disk commonly requires immediate surgery.
Ancillary Procedures: Radiographs of the lumbar and sacral vertebrae were normal. A myelogram and a computed tomography (CT) scan with a myelogram (see Video 10-1) showed a mild intramedullary swelling of the lumbosacral intumescence. The CT images are from the body of the L3 vertebra through the body of the L5 vertebra.
Because of the guarded prognosis, the owner of Brittney requested euthanasia. Necropsy confirmed extensive infarction scattered through the caudal lumbar and cranial sacral spinal cord segments. The gray matter was more extensively affected on the left side. There were numerous fibrocartilaginous emboli in spinal cord blood vessels associated with these lesions.
FCEM is a spinal cord lesion that is common in dogs but is uncommon in other species of domestic animals.* It is most common in young adult dogs of the larger breeds but it can occur as young as 3 months of age and it is common in the miniature schnauzer, Labrador retriever, and boxer breeds, in our experience. The clinical signs are peracute in onset and usually stabilize within 24 hours. Rarely, clinical signs may progress for 48 hours. Following that, there is no further progression or there is improvement, depending on the degree of ischemia or infarction that has occurred. The source of the fibrocartilage is assumed to be the intervertebral disk that is undergoing degeneration. This embolic fibrocartilage has the same collagen type that is found in the nucleus pulposus. How this degenerate fibrocartilage gains access to the spinal cord vasculature remains speculative. These emboli are more common in small arteries but also can be found in veins. Arteriovenous anastomoses do occur in the blood supply of the spinal cord and have been implicated in the distribution of the emboli. Protrusion of degenerate disk material into the adjacent ventral internal vertebral venous plexus has occasionally been observed at necropsy. It has been suggested that the normally avascular intervertebral disk is invaded by new growth of arteries when degeneration occurs in the annulus fibrosis and this is a route for these emboli to enter the arterial vasculature. We find this mechanism difficult to accept. In humans, degenerate intervertebral disk material can protrude into the adjacent vertebral body where there is ready access to the blood vessels in the marrow of the vertebra. One route of venous drainage from this marrow is into the ventral internal vertebral venous plexus within the vertebral canal. These intramedullary protrusions are referred to as Schmorl nodes. They are rare, or at least rarely identified in dogs, which may be because of dogs’ quadruped posture and the thick layer of cortical bone that is adjacent to the intervertebral disk. Reverse venous blood flow may be involved in the distribution of these emboli. Whenever an animal strains by contracting its trunk muscles with the glottis closed, the increased pressure in the thorax and abdomen interferes with the venous return to the heart and forces the venous blood into the vertebral venous plexus. This is the Valsalva maneuver, and it may play a role in the ability of these emboli to gain access to the spinal cord vasculature. The involvement of the intervertebral disk as the source of these emboli is also supported by the observation that these lesions occur primarily in the spinal cord. One report of brainstem lesions with fibrocartilaginous emboli indicated a possible source of emboli from cervical intervertebral disks.5 Magnetic resonance (MR) imaging often shows intervertebral disk degeneration at the level of the FCEM lesion in the spinal cord. These FCEM lesions can be unilateral or bilateral at any level of the spinal cord, and they affect various combinations of the gray and white matter. The lesions are usually limited to a few adjacent spinal cord segments. There are many examples in the following case examples that involve the various regions of the spinal cord. Caudal brainstem signs are rare and probably are associated with emboli arising from the cervical intervertebral disks.
Because of the extensive collateral circulation to the spinal cord (Fig. 10-4), multiple blood vessels must be compromised to cause the degree of infarction and severe clinical signs seen in dogs similar to Brittney. This suggests that a sudden shower of emboli must occur at one time. At necropsy, these emboli can be found in many blood vessels in or near the lesions. Usually, this shower affects the blood vessels to a few adjacent spinal cord segments and the associated lesions often are scattered and asymmetric within these segments. Thus, the clinical signs are usually focal and often asymmetric, as seen in Brittney. FCEM may be much more common than we realize and not be extensive enough to cause clinical signs or cause only transient clinical signs. Most veterinarians have had the experience of being called by a distraught owner who has just found their pet dog collapsed and unable to stand, but by the time the dog arrives at the hospital for examination, the dog is walking normally. We believe that some of these transient episodes of collapse may be due to transient spinal cord ischemia caused by FCEM.
Many dogs in which you make this clinical diagnosis will recover spontaneously. This is more common in dogs with paresis and ataxia due to interruption of the UMN and GP pathways. The more severe the involvement of the gray matter in the intumescences, the more guarded the prognosis. As a rule of thumb, if there are no signs of improvement in 10 to 14 days after the onset of clinical signs, it is unlikely that any recovery will occur. In a few patients, after an initial mild improvement, there may be a period of weeks before they rapidly regain the ability to walk. We usually tell owners that it may take up to 10 weeks before final improvement occurs.
An interesting observation is that FCEM is very rare in the chondrodystrophic breeds in which the chondroid metaplastic form of intervertebral disk degeneration is so common. FCEM also is observed in dogs as young as 3 months of age in which you do not expect intervertebral disk degeneration to occur. In these young dogs, the source of cartilage may be the vertebral growth plates. Often, there is an associated history of mild trauma such as a sudden fall or vigorous playing and jumping as with catching a frisbee. This relationship between young age and vigorous handling is associated with FCEM in young feeder pigs; it occurs during their transportation in crowded trucks.
The three primary disorders that can cause an acute onset of relatively nonprogressive spinal cord dysfunction are external injury by objects in the environment, most commonly vehicles; internal injury resulting from intervertebral disk extrusions; and vascular compromise resulting from FCEM. The history usually permits substantiation or exclusion of external injury. Lacking that, vertebral column radiographs should provide that answer. To differentiate between the other two causes of these clinical signs, evidence of discomfort by the patient is more suggestive of an intervertebral disk extrusion than of FCEM, but exceptions are common for both of these disorders. Ultimately, immediate imaging is necessary because a diagnosis of an intervertebral disk extrusion usually requires emergency surgery. Myelograms in dogs with FCEM are helpful only in the small percentage of dogs in which intramedullary swelling is extensive. MR imaging is much more reliable in detecting the spinal cord edema that accompanies the ischemia or infarction caused by the fibrocartilaginous emboli.25 Be aware that MR imaging that is done in the first 24 to 48 hours after the embolic shower occurs may occasionally be normal.
Examination: Video 10-2 was made 12 hours after the onset of the pelvic limb dysfunction. Note the paraplegia with thoracic limb hyperextension but normal voluntary movements in the thoracic limbs. Note the pronounced pelvic limb hypotonia but intact spinal reflexes. Note the line of analgesia at about the thoracolumbar junction.
The owner’s history of this dog’s clinical signs is sufficient to deny any external injury. It was difficult to determine how much of this dog’s struggling was due to discomfort, frustration, or dislike of the examiners. An acute intervertebral disk extrusion is less likely than FCEM in a 3-year-old Labrador retriever, due to its young age.
No ancillary studies were performed in Brandy, and with no clinical signs of improvement after a few days, the owners elected euthanasia. Necropsy revealed extensive ischemic and hemorrhagic infarction in the caudal thoracic spinal cord segments, with numerous fibrocartilaginous emboli in both arteries and veins. No intervertebral disk material was found on opening the ventral internal vertebral venous plexus in the area of the involved spinal cord segments, and a median plane section of the vertebral column did not reveal any recognizable intervertebral disk material in the marrow of the vertebral bodies.
Video 10-3 shows Spanky, an 11-year-old male mixed-breed dog that was riding in a car that skidded on ice and struck a guard rail. Radiographs (shown in the video) revealed a fracture in the spinous process of the T6 vertebra and a collapsed intervertebral disk space between the T5 and T6 vertebrae, with no displacement. The video was made 4 days after the injury. Note the pelvic limb hypertonia and hyperreflexia and therefore no evidence of any spinal shock. Note the presence of nociception in the pelvic limbs, indicating that the spinal cord lesion has not caused a complete transverse dysfunction. A body cast was applied and Spanky regained the ability to walk with his pelvic limbs in about 5 weeks.72
Video 10-4 shows a 6-year-old Parson (Jack) Russell terrier that was struck by an automobile and suffered a moderately displaced fracture of the L3 vertebra. The video shows the radiographs. Note the clinical signs of Schiff-Sherrington syndrome but the absence of any spinal shock.
In our experience, Schiff-Sherrington syndrome and spinal shock are rare in cats. The following three videos show cats with thoracolumbar vertebral column fractures that caused a transverse spinal cord dysfunction with peracute paraplegia and analgesia caudal to the lesion.
Video 10-5 shows Poison, an 8-year-old castrated male domestic shorthair that was struck by a vehicle 2 days before the video was made. Note the crossed extension in the pelvic limbs and the pelvic limb flexion when the tail was compressed with forceps. The latter is referred to as a mass reflex, which is another manifestation of hyperreflexia. Radiographs (see the video) showed a fracture at the articulation between the L3 and L4 vertebrae, with at least 50% displacement. Poison was euthanized and the necropsy showed that at the site of the fracture and displacement, there was only a sleeve of dura remaining. The parenchyma of the spinal cord at this site had been completely crushed and displaced into the adjacent spinal cord segments (Fig. 10-5).
Figure 10-5 Lumbar spinal cord of the cat in Video 10-5, showing a crease in the dura at the level of the fracture site. At this site there is only a sleeve of dura, and the parenchyma has been entirely displaced into the adjacent spinal cord segments.
Video 10-6 shows George, an adult Siamese cat that was found beside the road unable to use his pelvic limbs. Radiographs (see the video) showed a fracture at the articulation between the T13 and L1 vertebrae, with slight displacement. Note the pelvic limb flexion when the tail was compressed with forceps. This is an example of a mass reflex. George was euthanized and necropsy revealed hemorrhage and necrosis that involved all of the transverse section of the spinal cord at the level of the fracture (Fig. 10-6).
The spinal cord lesions resulting from injury caused by external trauma include the physical disruption of the high-risk spinal cord parenchyma as compared with the more resistant dural layer of meninges. See Fig. 10-6, which shows this severe spinal cord lesion. A contused spinal cord exhibits hemorrhage, edema, and necrosis due to compromise of the spinal cord’s vasculature. Following the traumatic event, there is continued degeneration of the injured spinal cord that progresses for a few hours, usually less than 24 hours. The basis for this is the subject of intense research directed at determining ways to treat these patients to prevent this progressive spinal cord destruction. Areas of interest include the release of neurotransmitters, such as toxic levels of glutamate, excessive accumulation of calcium ions, free radical species, nitrous oxide, and the release of various amines that cause vasoconstriction and subsequent ischemia and infarction.
Video 10-7 shows an 8-week-old Siamese kitten that was found beside the road by Society for the Prevention of Cruelty to Animals employees. Radiographs (see the video) showed a fracture at the T8-T9 vertebral articulation, with complete displacement through the vertebral canal. No owner was located. Despite the peracute paraplegia and analgesia caudal to the lesion, there were no signs of the Schiff-Sherrington syndrome or spinal shock. A medicine resident adopted the kitten, knowing that there would be no improvement in the clinical signs. The fracture was reduced and stabilized and a cart was made for the kitten (see the video). When the resident took a permanent position in Alaska, the then young adult cat was doing fine and enjoying life in a larger cart.
History: After a 5-mile run with his owners, Drake was in their home when he suddenly lost the ability to move his left pelvic limb. He was examined by a local veterinarian and then referred to Cornell University, where his disorder was filmed on the second day of the dysfunction, which had not changed since the onset.
Do not confuse this anatomic diagnosis with sciatic nerve paralysis, which looks much different. Review Video 5-44 of the collie dog Keane, described with Case Example 5-15; Keane has a left sciatic nerve injury due to a pelvic fracture. Dogs with sciatic nerve deficits can always advance the limb by hip flexion, and there is hypotonia of the crural muscles.
External injury was excluded based on the history and on the knowledge that such localized unilateral signs would be very unusual for an external spinal cord injury. The lack of any discomfort at the onset of the paralysis or during the examination is more suggestive of FCEM than of an intervertebral disk extrusion. An intervertebral disk extrusion would be unlikely in a 3-year-old nonchondrodystrophic breed. However, it can best be ruled down by imaging. Radiographs and a myelogram were normal, which supported a presumptive FCEM lesion.
On the video, you can follow the spontaneous recovery. The section of the video where Drake is walking indoors after you see his vigorous response to a noxious stimulus was made 1 week later, and the section outdoors was made 3 weeks after the initial filming.
Examination: This examination took place the day after the onset of the clinical signs, which had not changed. Video 10-9 shows that this dog’s clinical signs are the mirror image of those observed in Drake in Case Example 10-3. Penfield exhibits a spastic monoplegia of his right pelvic limb.
Differential Diagnosis: The diagnosis is the same as that described for Drake in Case Example 10-3. The indication of discomfort when the initial clinical signs occurred is suggestive of a possible intervertebral disk extrusion. Radiographs and a myelogram diagnosed a unilateral right-side intervertebral disk extrusion between the L3 and L4 vertebrae. This was removed via a hemilaminectomy, and Penfield was walking with the right pelvic limb after about 3 weeks.
Video 10-10 shows Dude, a 3-year-old male miniature schnauzer that had been outdoors unconfined during the night and was found in the morning unable to use his left pelvic limb. From your study of this video you should make the anatomic diagnosis of a focal spinal cord lesion on the left side between the T3 and L3 spinal cord segments. The differential diagnosis is the same as in this case example and Case Example 10-3 but must include external injury because the history cannot rule it out. The absence of any reluctance to try to walk as well as of any discomfort when palpated and the presence of extensive unilateral signs all suggest that external injury is unlikely. The young age of this dog suggests that a unilateral intervertebral disk extrusion is also less likely. In addition, this breed is at some risk for the development of fibrocartilaginous emboli. Radiographs and a myelogram were normal, which made FCEM the most presumptive diagnosis. Within 2 to 3 weeks, without specific therapy, Dude was walking well in the left pelvic limb. Figure 10-7 shows the kind of FCEM lesion that explains these clinical signs but from which recovery would not occur.
Figure 10-7 Transverse section of the T13 spinal cord segment of a 3-year-old Saint Bernard with a left pelvic limb spastic monoplegia due to a fibrocartilaginous embolism that caused ischemic infarction of the left half of the spinal cord. A similar anatomic but less severe lesion was presumed to have occurred in Drake in Video 10-8 and in Dude in Video 10-10.
Video 10-11 shows Beaver Dam, an 8-year-old spayed female mixed-breed dog that was presented because of difficulty using the left pelvic limb. This dysfunction began when the dog fell down a flight of stairs and had not changed significantly since that time. From your study of the video, you should make the same anatomic and differential diagnosis as for Dude in the previous video. The video of Beaver Dam shows the radiographs, myelogram, and CT imaging, which diagnosed intervertebral disk extrusions at the L1-L2, and L3-L4 vertebral articulations. The owner elected to treat the dog medically, and no follow-up on its success is available.
Paraparesis and Ataxia
History: About 4 weeks before examination at Cornell University, this dog was on a camping trip in the Adirondacks when the owners first noticed that Cassidy occasionally stumbled with his left pelvic limb. Within the next week he was also stumbling with the right pelvic limb. This pelvic limb dysfunction slowly progressed until the time of this examination.
Examination: Video 10-12 shows the monoplegia of the left pelvic limb and grade 2 monoparesis and ataxia of the right pelvic limb, with absent postural reactions. Cranial nerves and thoracic limb function were all normal. Note the mild hypertonia and normal spinal reflexes for the pelvic limbs and the normal nociception.
With normal nociception, there is no reliable way to localize a focal lesion between these segments using the physical neurologic examination. Be aware that there could be multiple or diffuse lesions in this anatomic area that would explain the clinical signs that were observed in this dog.
Differential Diagnosis: The following diagnoses are limited to those spinal cord disorders that could cause a progressive lesion within these thoracolumbar spinal cord segments of a dog: neoplasm, vertebral malformation, myelitis, discospondylitis, multiple cartilaginous exostosis, intervertebral disk extrusion-protrusion, degenerative myelopathy.
Although we tend to relegate neoplasia to older dogs, there are many exceptions to this, and one of them occurs at this location in dogs. Nephroblastoma is a unique intradural, extramedullary (extraaxial) neoplasm that occurs in young dogs, usually between spinal cord segments T10 and L2 (Figs. 10-8 through 10-10).91 Most of our experience and reports in the literature involve dogs younger than 3.5 years of age, many of them only a few months old. There are numerous reports, mostly in the European literature, that describe this neoplasm as an ependymoma.58,92,109 In the first two editions of this text, I (AD) recognized that this was not intramedullary and therefore was not an ependymoma, and I called it a neuroepithelioma based on the morphology of the cells and the abundance of tubular elements in the neoplasm. Since then, we have recognized tubular structures that resemble renal glomeruli and have diagnosed this neoplasm as a presumptive nephroblastoma. It is an embryonic neoplasm that may arise from mesonephric tubules. This is supported by the positive immunocytochemical staining of the neoplastic cells for a polysialic acid marker of this tumor; the test was developed in children by Dr. J. Roth in Switzerland.78
In children, this nephroblastoma occurs in the kidney and is called a Wilm tumor. A gene located on chromosome 11 has been identified for this tumor in children. Using antibodies developed for the protein product of this tumor gene, immunocytochemical staining has identified this protein product in the canine neoplasm.76 The unique location of this neoplasm between spinal cord segments T10 and L2 correlates with the site of embryonic renal development from intermediate mesoderm. This is adjacent to the development of the somitic sclerotomes that envelope the neural tube that forms the thoracolumbar spinal cord. A Wilm tumor has three components, sheets of unorganized epithelial cells referred to as blastemal cells; tubular elements lined by epithelial cells that vary from squamous to cuboidal and some of which form glomerular structures; and a fibrous component that consists primarily of bundles of collagen. The canine neoplasm is made up primarily of the first two components (Figs. 10-11 through 10-15). A nephroblastoma in this spinal cord location is very rare in children. In dogs, this neoplasm does occur in the kidney but it is much more common adjacent to the spinal cord. There are no reports of this neoplasm being in both locations in the same patient.
Typically, the clinical signs occur in one pelvic limb prior to their occurrence in the other, which reflects the asymmetric location of the neoplasm. We suspect this is a very slowly growing neoplasm and recognize that the spinal cord can be slowly compressed by an extramedullary mass for a long time before the autoregulation of spinal cord vascular perfusion fails and clinical signs occur. Therefore, when these clinical signs occur, be aware that the spinal cord is already very compressed, and if surgery is to be considered it should be done as soon as possible. Often at necropsy, the spinal cord consists only of a thin 1- to 3-mm quarter-moon-shaped shell covering one side of the neoplasm (see Figs. 10-10 through 10-12). When surgeons remove this neoplasm they often see the severely compressed spinal cord slowly start to fill the space where the mass was removed. Postoperative radiation therapy should be done to avoid recurrence of the neoplasm.
Another neoplasm that is more common in cats and can occur at any level of the vertebral canal is lymphoma. In cats it commonly occurs before 1 year of age. Gliomas are less common in the spinal cord of all species and usually occur in older animals.
Vertebral malformation with kyphosis and secondary spinal cord compression is a realistic consideration in this dog. This vertebral malformation is usually in the midthoracic portion of the vertebral column.49,67 Although these dogs have the vertebral malformation at birth, the clinical signs of spinal cord compression usually do not occur until a few months of age but prior to 1 year of age. We believe that the kyphosis that causes the spinal cord compression develops at the site of the malformation as the dog grows, and this accounts for the age of onset of the progressive spastic paraparesis and ataxia of the pelvic limbs. Occasionally, on your physical examination of the patient, you can see and palpate the kyphosis. However, it is easy to overlook if you do not take the time to carefully examine the patient for it. This malformation is readily seen on radiographs. It was not visible or palpated in Cassidy. (See Videos 10-15 and 10-16.)