N. Fitzpatrick
Degenerative lumbosacral stenosis (DLSS) is the most common cause of compression of the cauda equina and seventh lumbar (L7) nerve roots in dogs. DLSS is characterised by lumbosacral (LS) intervertebral disc (IVD) protrusion, subluxation or inflammation of the facet joints with associated thickening of the joint capsule and hypertrophy of the interarcuate ligament. The lumbosacral joint is the most mobile functional spinal unit (FSU)in dogs. Pathological static or dynamic alteration in load transmission across the L7-S1 joint is thought to be the most common contributor to lumbosacral intervertebral disc degeneration in dogs. A common sequel to disc-associated DLSS is impingement of the nerve roots or vasculature of the cauda equina and/or L7 nerve roots, either by the disc itself or by secondary inflammatory, fibrous, or osseous impingement. Abaxial disc protrusion and spondylosis can be a significant cause of pain and lameness in large breed dogs whilst lower motor neuron deficits are much less common in our case population in the UK.
Key clinical features are unwillingness to jump, a hunched stance and uni- or bi-lateral pelvic limb lameness, which may occur after prolonged recumbency and lessen with exercise. The sciatic nerve pathway can be palpated per rectum and in the recess of the caudal thigh musculature, whereupon deep digital pressure may produce resentment and vocalisation, though interpretation of this test is subjective and experience in normal and diseased dogs is important. Focal pressure application to the dorsal aspect of the lumbo-sacral junction is also important albeit subjective and hyperextension of the lumbosacral junction in isolation without caudal extension of the coxofemoral joints is a useful interrogation during clinical examination. Objective measures of lumbosacral function and pain using kinematics and electrophysiology are evolving and may yield greater accuracy with regard to detection of clinically relevant pain. It is critical to exclude coxofemoral and stifle disease as well as other causes of pelvic limb pain and lameness and infections or neoplastic pathologies must always be considered.
Diagnosis is generally based on clinical examination and advanced imaging. MRI has superior soft tissue contrast resolution and sensitivity for detection of IVD degeneration is high. However, MRI lacks the ability to provide a reliable correlation between severity of clinical signs and the severity of the compression. CT is more valuable for assessment of definitive osseous boundaries, but good agreement between MRI and CT findings has been documented. MRI is frequently used in canine and human patients to identify primary lumbar foraminal stenosis, since unrecognised or recurrent foraminal stenosis may be associated with “failed back surgery syndrome.” L7-S1 foraminal stenosis and associated compressive radiculopathy has been documented to occur with a reported incidence of 68–90% in dogs presenting with clinical signs associated with DLSS. MRI interrogation of the LS spine in dogs has been used to divide the L7-S1 intervertebral foramen (IVF) into entrance, middle and exit zones, allowing description of a foramin as stenotic when loss of the fat signal is complete or when only a minimal rim of fat signal is visible within one of the foraminal zones on parasagittal imaging plane sequences.
It is important to recognise that for some patients, especially working and agility dogs, dynamic imaging is fundamental to diagnosis. In the author’s practice we see a significant number of canine athletes that present with performance impairment attributable to dynamic claudication of the cauda equina and L7 nerve roots. We routinely image lumbosacral spinal motion in neutral and hyperextension and have shown that neuroforaminal dimension varies considerably and that this is clinically relevant in cases of intervertebral disc protrusion with or without new bone formation. Additionally, the dorsal lamina of the sacrum and articular facet proliferation may encroach dorsally on both the sacral and the L7 nerve roots. These changes may not be apparent on neutral sagittal, parasagittal, and transverse imaging planes.
Acquisition of MRI images in conventional parasagittal planes may fail to accurately represent the actual dimensions of the entry, middle, and exit zones of the neuroforamina. We have performed both a cadaveric and an in vivo comparison of standard parasagittal plane image acquisition with parasagittal oblique imaging where images are obtained perpendicular to the L7 nerve pathway rather than parallel to the sagittal plane. These studies have revealed greater sensitivity of oblique parasagittal imaging for detection of encroachment of the L7 nerve pathway in all zones.
Treatment methods for DLSS have been broadly divided into non-surgical and surgical, with the recommended treatment depending on the nature of cauda equina and nerve root compression (static or dynamic, central or lateralised, acute or chronic) the severity and chronicity of clinical signs and the intended use of the dog. Our experience with epidural steroid injections has been favourable for dogs affected by moderate encroachment of the L7 neuroforamina and pain only.
Although the primary aim of surgical intervention has been cited as acute cauda equina and neuroforaminal decompression, long-term maintenance of decompression and amelioration of pathological LS instability are important secondary aims. Dorsal decompressive procedures are known to destabilise the lumbar vertebral column, particularly where facetectomy is performed. While lateral foraminotomy is likely to cause minimal additional instability unless the facet is undermined, in which case fracture could occur, it may also be challenging to decompress the nerve root adequately and iatrogenic trauma could occur. It may be necessary to combine a lateral and a dorsal approach to attempt to decompress entry, middle, and exit zones. Durability of decompression is also unknown, with osseous or soft tissue re-growth remaining a potential concern, either via post-surgical scarring or in association with recurrent peri-vertebral soft tissue or osseous proliferation as a result of ongoing instability.
Foraminotomy, facetectomy, or dorsal stabilisation without distraction (using pins and cement, screws alone, or plates and screws) may, therefore, fail to provide durable resolution of L7 nerve root encroachment in the entry, middle, and exit zones. Our aim over the last decade was to develop an intervertebral spacer screw that could be deployed via dorsal laminectomy and dorsal annulotomy with nuclear extirpation following lateral retraction of the cauda equina. Dorsal fixation elements would then be applied using pins or screws and formerly cement but more recently, rods and clamps (Fitzateur)to produce a dorsal and ventral fixation system akin to anterior-posterior inter-body fusion performed in human patients.
The spacer device has been called a Fitz intervertebral traction screw (FITS). This constitutes a threaded titanium spacer screw with hydroxyapatite coating and is currently available in ten sizes. A 2.4 mm screw is driven cranio-ventrally from S1 to L7 through a slot in the FITS device to prevent back-out. The dorsal fixation system is currently available in two sizes. Either 3.5 mm or 4.5 mm screws are placed on both sides of the L7 vertebra at the bases of the transverse processes and in the alar wings of the sacrum on both sides. A novel slotted polyaxial screw fixation clamp system, which allows multidirectional orientation, permits the L7 and sacral screws to be linked dorsally using rods. The rods are dumb-bell shaped to firmly anchor in the clamps to minimise risk of collapse. The clamps are locked to the spherical stopper ends of the rods using washers and lock-screws.
It should be noted that retraction of the cauda equina and stretching of chronic compressive elements away from the L7 nerve roots is not a benign intervention and risks should be explained to all owners pre-operatively. It should also be noted that other fusion systems such as custom-designed 3D printed plates and screw-rods designed for human patients have been successfully deployed in other centres. It is likely that other surgical techniques will also be developed for treatment of this difficult condition and they may be equal or superior to the system described.
In a recent case series presented by our facility, thirty-one out of thirty-five patients exhibited improvement in activity and lameness. Twenty-nine out of thirty-five owners felt quality of life had improved after intervention and two out of thirty-five felt it was equal to pre-surgical levels. Four out of thirty-five owners were indifferent due to complications or failure to improve. All patients were receiving NSAIDs at presentation whereas twenty-five out of thirty-five did not require NSAID therapy postoperatively. Veterinary examination was available for thirty-four out of thirty-five patients. Lameness, pain, and muscle mass scores all improved. On CT scans, significant difference was found between pre-operative and six months postoperative measurements of all endplate distances in sagittal and dorsal planes and also all neuroforaminal dimensions in all zones on both sides. Bone in-growth was observed around the FITS device.
We have performed mechanical testing of the FITSFitzateur system wherein instrumented lumbosacral spinal units were subjected to non-destructive 4-point bending under compressive loading. Dorsal laminectomy/discectomy resulted in a modest increase in the range of motion at L6-L7 and L7-S1 as compared with the intact spine when subjected to compressive loading up to 150N. Application of the spinal instrumentation at L7-S1 resulted in a significant reduction in flexion, extension, and lateral bending at L7-S1 as compared with laminectomy alone, but no significant change in motion at the L6-L7 junction.
The FITS-Fitzateur system allows indirect decompression of the neuroforamina and provides durable long-term stability. This may reduce the propensity for “failed back syndrome” whereby failure to decompress the components of abaxial encroachment may result in ongoing pain which can be due to static or dynamic compression. Facet joint capsule thickening and osteophyte formation are also potentially addressed by distraction and stabilisation.
In conclusion, in dogs affected by DLSS, conservative management may not be effective in relieving pain and surgical intervention may be needed. Dorsal laminectomy and facetectomy with or without partial discectomy results in increased motion. Foraminotomy may not produce durable neuroforaminal decompression. Pedicle-screw systems have been shown to be effective in restoring stability of the LS junction but traditional mono-axial screws lack versatility. Instrumentation with the novel spinal system described results in statistically significant reductions in lumbosacral instability in extension, flexion and lateral bending. The new instrumentation also has two important potential mechanical advantages over standard pedicle screw-rod combinations. First, use of a polyaxial head on the screw allows for far greater versatility when linking the screw to the connecting rod (Fitzateur). Second, deployment of an interbody spacer (FITS) permits distraction, thereby restoring the neuroforamen and providing a conduit for fibro-osseous fusion of the interbody space. In clinical cases, the Fitzateur also provides more soft tissue cover and superior implant-bone purchase than the predecessor technique using pins and cement.
It is clear that more objective tests to define clinical severity of DLSS need to be developed based on functional questionnaires, electrodiagnostics, kinetic and kinematic profiles, and similarly more objective outcome measures following intervention. Dogs suffering pain and disability associated with DLSS can be treated with oral medication, epidural steroid, or surgical intervention, depending on the patient, on degree of compression and on severity of clinical signs. Randomised comparative trials would require large treatment cohorts to provide adequate power to discern between treatment differences and, until clinical and imaging inclusion/exclusion criteria are established, this will be challenging.
There are very different sub-groups of clinical cases varying from chronic osseous, discogenic and inflammatory impingement which is apparently adequately tolerated by sedentary dogs through to mild dynamic impingement poorly tolerated by very athletic dogs. Furthermore, it is difficult to conduct unbiased randomised comparative trials with client-owned dogs. Therefore, it behoves us to focus on more rigorous case definition, developing better understanding of the correlation between clinical signs and imaging findings and being as critical and transparent as possible in outcomes analysis.