Three skeletally immature screw-tailed dogs were evaluated because of progressive pelvic limb ataxia. Dog 1 was a 6-month-old sexually intact male Pug, dog 2 was a 7-month-old sexually intact male English Bulldog, and dog 3 was a 4-month-old sexually intact male Pug. In all dogs, findings on physical examination were unremarkable, with the exception of palpable deviation of the normal contour of the thoracic vertebral column despite apparently normal epaxial muscle bulk. At the time of initial evaluation, all the dogs were ambulatory, although marked pelvic limb ataxia and paresis were evident; the condition of dog 2 deteriorated rapidly and it developed nonambulatory paraparesis within 24 hours of admission to the hospital. Neurologic examination of each dog revealed apparently normal function in the thoracic limbs; in the pelvic limbs, flexor and patellar reflexes were intact but there were deficits in responses to toe knuckling and hopping. Dog 1 had a crossed extensor reflex (from right to left) in the pelvic limbs. The panniculus reflex could not be elicited in dog 1, was weak caudal to the level of the T8 vertebra in dog 2, and was considered normal in dog 3. Signs of mild discomfort were evoked on forcible palpation of the midthoracic region of the vertebral column in each dog. These findings suggested that there was a lesion between the T3 and L3 spinal cord segments in all 3 dogs.
Radiography was performed on each dog and revealed marked kyphosis of the thoracic vertebral column that was centered on the region where there was 1 or more hemivertebrae. In dog 1 (the 6-month-old Pug), this region was centered on the T9 vertebra; the vertebral body was notably short in the craniocaudal plane and displaced dorsally with respect to the neighboring T8 vertebra. The T8 vertebral body was also shortened, and there was bone proliferation on the ventral aspect of the T8-9 intervertebral space. Ventrodorsal radiographic views of dog 1 did not reveal any lateral displacement of the affected or neighboring vertebrae. In dog 2 (the 7-month-old English Bulldog), there was abnormal curvature of the thoracic portion of the vertebral column that was centered on the T6-7 intervertebral junction; the T5, T6, and T7 vertebral bodies were markedly short craniocaudally. The T8 vertebra was wedge shaped, and the apex was directed dorsally (Figure 1). Ventrodorsal radiographic views of this region revealed an S-shaped deformity of the vertebral column. Several vertebrae at this level appeared to be incompletely developed and, in some instances, butterfly shaped. In dog 3 (the 4-month-old Pug), the T8 vertebra was trapezoid shaped on lateral radiographic views (the narrow edge was directed cranially) and malarticulated and dorsally displaced with respect to the T7 vertebra.
Lateral radiographic views of 2 dogs with hemivertebrae. A—Lateral radiographic view of dog 2. Notice the triangular shape of the T8 vertebral body (apex dorsal). The vertebral body of T7 is malarticulated and dorsally displaced (arrow) with respect to T8, suggestive of compression of the spinal cord at this site. In this dog, 2 ribs attached to T8 appear to be fused (asterisk). B—Lateral radiographic view of dog 3. Notice the kyphosis centered on theT8 vertebra.There is apparent loss of the cranial growth plate ofT8 and malarticulation with T7. Marked deformity of the laminae and spinous processes of T6, T7, and T8 is also apparent.
Citation: Journal of the American Veterinary Medical Association 230, 4; 10.2460/javma.230.4.532
For each dog, T1-weighted (time to repeat, 600 to 800 ms; time to echo, 26 ms) and T2-weighted (time to repeat, 3,000 ms; time to echo, 80 ms) MR images were obtained by use of a 0.2-T permanent magneta (Figure 2). Sagittal MR images of dog 1 suggested that there was severe spinal cord compression associated with the region of severe kyphosis and the malformation and dorsal displacement of the T9 vertebral body. Transverse MR images confirmed that there was severe ventrodorsal spinal cord compression associated with gross reduction in the dorsoventral height of the pedicles (compared with findings expected in clinically normal dogs) at this site. The region of compression was only approximately 6 mm long. These images also revealed a component of ventrolateral cord compression that was apparently associated with the imperfectly formed dorsal aspect of the T9 vertebral body.
Magnetic resonance images of 3 dogs with hemivertebrae. A—Midsagittal T1-weighted image of dog 1. Notice the severe spinal cord compression associated with hemivertebra atT9. B—TransverseT2-weighted image of dog 1 at the level indicated by the dashed line in panel A. Notice the severe stenosis of the vertebral canal, which is associated with grossly shortened pedicles (arrow). The spinal cord is hyperintense, compared with regions of normal spinal cord in adjacent sections (not shown), which is suggestive of edema. C—Midsagittal T1-weighted image of dog 2. Notice that the spinal cord compression appears relatively mild, although the spinal cord is partially out of the plane of section because of the lateral curvature of the vertebral column in this dog. D—Transverse T1-weighted image of dog 2 at the level indicated by the dashed line in panel C. Notice that the ventrolateral spinal cord compression is severe (arrow). At surgery, cancellous bone was identified at the site of compression. E—Midsagittal T2-weighted image of dog 3. Notice that the spinal cord compression is associated with the hemivertebra at T8. F—Transverse T2-weighted image of dog 3 at the level indicated by the dashed line in panel E. Findings in this dog are similar to those of dog 1 (also a Pug); gross stenosis of the vertebral canal is present as a result of shortening of the pedicles (arrow) and hyperintensity of the cord.
Citation: Journal of the American Veterinary Medical Association 230, 4; 10.2460/javma.230.4.532
In dog 2, sagittal MR images confirmed that there were extensive abnormalities of the T4, T5, and T6 vertebral bodies, but because of the lateral deviation of the vertebral column at this level, it was difficult to define precisely the origin or degree of cord compression. Transverse MR images of this region confirmed that the vertebral canal was severely stenotic in both planes. The T1-weighted transverse scans indicated that there was lateral compression of the spinal cord associated with a mass to the right of the cord within the vertebral canal at T5; the mass was hyperintense, compared with both the spinal cord and the vertebral bone.
In dog 3, sagittal MR images confirmed the abnormal development and incomplete separation of T7 and T8 vertebrae (the intervertebral disk between these 2 vertebral bodies was absent) and suggested that this was associated with spinal cord compression. Transverse images indicated that there was severe dorsoventral attenuation of the vertebral canal over a distance of approximately 10 mm, which was associated with gross dorsoventral shortening of the pedicles.
Each dog was treated surgically. The precise surgical approach was determined by the origin and nature of the compressive lesion that was revealed via MR imaging. Thus, in dog 1, dorsal laminectomy over the T8, T9, and T10 vertebral bodies (where there was considerable ventrodorsal cord compression) was performed. In addition, the annuli of the T8-9 and T9-10 intervertebral disks were excised as completely as possible to allow the vertebral body of T9 to be displaced ventrally. Ventral displacement was accomplished through leverage obtained by implantation of positive-threaded external fixator pinsb into the vertebral bodies of T8, T9, and T10 (Figures 3 and 4). The vertebrae were then fixed in position by application of PMMA bone cementc (in a manner similar to its use for stabilization of vertebral fracture-dislocations1). In dog 2, a right-sided hemilaminectomy of T4, T5, and T6 vertebrae allowed resolution of laterally located compression caused by encroachment of a hypertrophic articular facet into the vertebral canal. The region was then stabilized by use of a unilateral construct of positive-threaded pinsb driven into the vertebral bodies and embedded in PMMA bone cement. In dog 3, a dorsal laminectomy was used to decompress the spinal cord within the T7 and T8 vertebrae, followed by segmental spinal stabilization2 with 1.0-mm-diameter Kirschner wires attached with 2-0 polypropylene sutured and PMMA bone cement (to prevent pin migration) to the spinous processes, articular facets, and rib heads from T3 through T12.
Lateral radiographic views of 2 dogs after surgical treatment of hemivertebrae. A—Radiographic view of dog 1. Threaded pins have been placed into T6, T7, and T8 vertebral bodies and stabilized with PMMA cement. The wire was used to aid in anchoring the PMMA to the pins. A similar construct, although placed unilaterally, was used in dog 2. B—Radiographic view of dog 3.The contour of the affected region of the vertebral column has been altered and stabilized by use of a segmental stabilization technique augmented by application of PMMA cement.
Citation: Journal of the American Veterinary Medical Association 230, 4; 10.2460/javma.230.4.532
Lateral radiographic views of dog 1 obtained before (A) and after (B) surgical treatment of a hemivertebra. A line has been superimposed to follow the ventral aspect of the vertebral canal in both images. Before surgery, the angle of kyphosis was 126°; after surgery, the angle of kyphosis was 132°.
Citation: Journal of the American Veterinary Medical Association 230, 4; 10.2460/javma.230.4.532
After surgery, neurologic recovery was slow in dog 1; after 2 months, the dog was mildly ataxic with persistent mild deficits in toe knuckling and hopping responses. This dog subsequently reached skeletal maturity and remained mildly ataxic in the pelvic limbs, despite only equivocal improvement in degree of kyphosis (Figure 3). In dog 2, there was early neurologic deterioration following surgery because of cord compression induced by bone cement (and an adjacent insulating gelatin spongee) encroaching on the hemilaminectomy site. Given the radiographic evidence of this, the dog underwent exploratory surgery during which the encroaching cement was identified and removed; new PMMA bone cement was placed more laterally to ensure that encroachment on the vertebral canal did not recur. The dog subsequently recovered and had mild ataxia and mild deficits in hopping responses by 12 weeks after surgery. Radiography revealed that the stabilized vertebrae maintained their immediate postoperative positions with respect to one another despite growth of neighboring vertebral bodies. Following surgery, dog 3 recovered rapidly and had a normal gait after 3 weeks; mild deficits in pelvic limb hopping responses were evident at that time and this level of function was maintained while growth was completed.
Discussion
Hemivertebrae are commonly detected in brachycephalic breeds of dog, most notably in those that also have screw tails, such as English Bulldogs, French Bulldogs, Boston Terriers, and Pugs.3,4 A hemivertebra is thought to result from asymmetrical development or fusion of the 2 ossification centers of the centrum of 1 or more vertebral bodies,5 which leads to various deformations of the vertebral bodies depending on the plane in which the malformation develops. Thus, there can be ventral, dorsal, or lateralized wedging of the vertebral body that often appears to cause responsive alterations in the shape of adjacent vertebral bodies. The high frequency of the condition in screw-tailed breeds of dog is to be expected because a screw tail is a consequence of a hemivertebra in the coccygeal region.3
The most commonly affected site is the midthoracic region of the vertebral column, particularly in association with the T8 vertebra,3 and although hemivertebrae are usually incidental findings, clinical signs (when present) are suggestive of transverse myelopathy that affects the T3 through L3 segments of the spinal cord. Typical signs are pelvic limb ataxia and paresis with intact segmental reflexes; severity can vary greatly, from very mild ataxia to complete loss of voluntary movement of the pelvic limbs and both fecal and urinary continence, although loss of pain perception is rare. Assessment of the panniculus reflex can aid in precise localization of the lesion, but frequently, the affected region of the vertebral column is apparent simply by palpation. Radiographically, the diagnosis of hemivertebra is straightforward, but it is important that the possibility of concomitant developmental spinal cord anomalies is not overlooked as a potential cause of the clinical signs3,6; therefore, application of myelography or advanced imaging techniques is required.
Because neurologic deficits associated with hemivertebrae in dogs are frequently mild and nonprogressive3 and stabilize as growth of the vertebrae stops (ie, at approx 9 months7), conservative treatment is often appropriate. Nevertheless, there are some affected dogs in which the clinical signs progress to cause unacceptable loss of function before completion of growth, and for those individuals, a surgical solution is required. Although hemivertebra formation is a well-recognized condition in dogs, there are no detailed reports of treatment to our knowledge, although surgery has been alluded to in previous publications.5,6,8
There have been several obstacles to achieving reliable surgical treatment of hemivertebrae in dogs. Until recently, there was considerable difficulty in determining the precise cause and location of compression of the spinal cord because of the limited information provided by radiography and myelography. The advent of widespread availability of advanced imaging techniques that are capable of providing transverse (axial) images, such as MR imaging and CT, for use in canine patients has largely removed this obstacle by allowing excellent 3-dimensional imaging.9 However, although CT and CT-assisted myelography have been widely used for improved refinement in diagnosis of many vertebral column−related conditions in dogs,10,11 application of either technique in the diagnosis of spinal cord compression associated with hemivertebrae has not been reported to our knowledge. As illustrated by the dogs of this report, MR imaging can reveal some hitherto difficult-to-appreciate features of the condition itself and detect congenital spinal cord malformations that have developed concurrently.3,6 Transverse images provide the most reliable data for diagnosis of spinal cord compression associated with hemivertebrae; the lateral deviation of the vertebral column that is common in this condition renders sagittal images unreliable because they may provide partially oblique views that include the epidural space. A notable component of the cord compression associated with hemivertebrae appears to be gross shortening of the pedicles; as a result, the spinal cord is enclosed within an aperture that is grossly narrowed dorsoventrally. This problem is then compounded by dorsal or lateral displacement of the affected vertebral bodies.
Also, there are technical difficulties in robust stabilization of bones that are soft and still growing, which is frequently a problem because the clinical signs usually develop in immature dogs. Furthermore, there is a lack of purpose-designed implants for treatment of vertebral column deformities in dogs, which contrasts sharply with the wide availability of such devices for treatment of humans with similar conditions. Iatrogenic induction of instability is a major concern in treatment of hemivertebrae. The trapezoid or triangular shape of the hemivertebral bodies implies that the force that is normally predominantly rostrocaudally directed along the vertebral column (as a consequence of the bowstring effect of the abdominal musculature12) will be more greatly diverted into a dorsally directed vector, which causes either affected or neighboring vertebral bodies to be displaced dorsally during bone growth.5 The MR imaging diagnosis of vertebral canal stenosis implies that dorsal laminectomy is often appropriate (eg, dogs 1 and 3 of this report), but removal of the dorsal stabilizing influence of the dorsal lamina could be hazardous because it will interrupt the integrity of 1 of the 3 so-called columns of support of the vertebral column.13 Therefore, an important aspect of treatment is to provide support through spinal fixation that enhances the strength of the dorsal column, such as segmental spinal stabilization2 or the combined pin and PMMA1 technique. It appears from the results obtained in the dogs of this report that suitable fixation can be achieved and might also reduce the degree of angulation in the affected region of the vertebral column, although achieving reduction in angulation is not straightforward. Overall, our experience with these dogs suggests that careful examination of MR images of dogs affected by hemivertebrae and tailoring the approach to spinal cord decompression and vertebral stabilization for each individual can lead to satisfactory outcomes.
ABBREVIATIONS
| MR | Magnetic resonance |
| PMMA | Polymethylmethacrylate |
| CT | Computed tomography |
Vet MRI, Esaote, Genova, Italy.
Veterinary Instrumentation, Sheffield, UK.
Palacos R-20 with Gentamicin, Schering-Plough Europe, Brussels, Belgium.
Prolene, Johnson & Johnson Medical Ltd, Gargrave, Skipton, UK.
Spongostan, Johnson & Johnson Medical Ltd, Gargrave, Skipton, UK.
References
- 1↑
Blass CE, Seim HB. Spinal fixation in dogs using Steinman pins and methylmethacrylate. Vet Surg 1984;13:203–210.
- 2↑
McAnulty JF, Lenehan TM, Maletz LM. Modfied segmental spinal stabilisation in repair of spinal fractures and luxations in dogs. Vet Surg 1986;15:143–149.
- 3↑
Done SH, Drew RA, Robins GM, et al. Hemivertebra in the dog: clinical and pathological observations. Vet Rec 1975;96:313–317.
- 4
Wright JA. Congenital and developmental abnormalities of the vertebrae. J Small Anim Pract 1979;20:625–634.
- 5↑
Colter SB. Congenital abnormalities of the spine. In:Bojrab MJ, ed.Disease mechanisms in small animal surgery. 2nd ed.Philadelphia: Lea & Febiger, 1993;950–959.
- 6
Bailey CS, Morgan JP. Congenital spinal malformations. Vet Clin North Am Small Anim Pract 1992;22:985–1015.
- 7↑
Walker MA. The vertebrae—canine and feline. In:Thrall DE, ed.Textbook of veterinary diagnostic radiology. 4th ed.Philadelphia: WB Saunders Co, 2002;98–109.
- 8
Sharp NJH, Wheeler SJ. Miscellaneous conditions. In:Sharp NJH, Wheeler SJ. Small animal spinal disorders: diagnosis and surgery. 2nd ed.London: Elsevier Mosby, 2005;319–337.
- 9↑
Bagley RS. Diagnostic testing in animals with spinal disease. In:Bagley RS. Fundamentals of veterinary clinical neurology. Ames, Iowa: Blackwell Publishing Professional, 2005;255–290.
- 10
Olby NJ, Munana KR, Sharp NJ, et al. The computed tomographic appearance of acute thoracolumbar intervertebral disc herniations in dogs. Vet Radiol Ultrasound 2000;41:396–402.
- 11
Sharp NJH, Wheeler SJ, Cofone M. Radiological evaluation of “wobbler” syndrome—caudal cervical spondylomyelopathy. J Small Anim Pract 1992;33:491–499.
- 12↑
Hansen H-J. A pathologic-anatomical study on disc degeneration in dog, with special reference to the so-called enchondrosis intervertebralis. Acta Ortho Scand Suppl 1952;11:1–117.
- 13↑
Shores A. Fractures and luxations of the vertebral column. Vet Clin North Am Small Anim Pract 1992;22:171–180.
