The translation of therapies from laboratory animals to human patients necessitates defining the dose of a given treatment required to obtain the expected effect in a different, much larger, species. Defining the dose may be a problem in itself although may be done by simply scaling up the rat dose for the body weight, or body surface area, of a human patient (which may be an adequate approach for drug administration). Another method, which may be a better approach for cell therapies, would be to scale up the dose for the size of the spinal cord or by correcting the dose for the volume of the lesion measured by magnetic resonance imaging (MRI) (Fig. 2) (36). For example, the normal spinal cord diameter of Dachshund dogs in the caudal thoracic region (T9) is 4.0  ±  0.5 mm and 3.0  ±  0.2 mm at a lesion site (T12, T13 or L1) 3 months after an injury (n  =  10 dogs with a mean body weight of 6.2  ±  1.8 kg; unpublished data; measurement from MR images). In comparison, the rat normal spinal cord diameter measured on MRI varies between 1 and 3 mm (36), and the gross human spinal cord diameter is around 7 mm in the caudal thoracic region (37).

However, because human or dog spinal cord lesions are longer and wider than lesions obtained in laboratory animals, not only the number of cells but the migratory properties of the cells tested, as well as the optimal method of delivery of the cells to populate the entire lesion must also be considered. The key difference between cell and drug therapies in this regard is that, although the calculated number of cells may be appropriate, it will also be critical that the cells are able to arrive at a lesion site within a specified amount of time (i.e., when the environment is responsive to their effects). The size of a lesion may present an obstacle in this respect because the time that is required for a cell to migrate through a large lesion may delay it sufficiently to make it arrive too late to be of benefit. Therefore, the dog forms a good model in which treatment doses and timing could be tested for their flexibility in development toward human use, since spinal cord lesions are larger than those in rats.

The availability of human spinal cord pathological specimens is very limited and when available, poor tissue preservation due to delayed fixation can limit acquisition of high quality data. Dogs have a shorter lifespan that humans, and may live only between a few months to several years following SCI and so are more commonly available for post mortem examination including, in some cases, tissue prepared by perfusion-fixation. Owners of dogs that have taken part in exploratory studies on novel therapeutic interventions are highly motivated by the research and commonly allow post mortem examination.

Recovery following a SCI is highly influenced by the activities of the patient, either performed autonomously (if still possible) or via a rehabilitation program. For example, treadmill training can have a major impact on plasticity of the intrinsic spinal cord circuitry in humans, cats and rats (38). It has also recently been found that behaviors can compete. For example, environmental enrichment in rats (which favors non-specific “self-training”) can cancel the benefits obtained by combining an intervention with specific task training (39). Following SCI, domestic dogs are treated in directly comparable environments as human patients and may also benefit from rehabilitation, such as physiotherapy and hydrotherapy, in which dogs engage very readily. Thus, domestic dogs could also be used to investigate how best to integrate medical interventions and rehabilitation programs.

All types of lesion found in humans, including pure contusion, mixed contusion-compression, chronic compression, dynamic compression and vascular lesions have been recognized in dogs, and are readily identifiable on advanced imaging studies. However, penetrating trauma, such as that resulting from knife or gunshot wounds, is uncommon.

Domestic dogs are therefore available for modeling many different aspects of SCI although, as in humans, the most common clinical lesion results from acute mixed compressive and contusive events, such as that associated with fracture-luxation or intervertebral disk nuclear extrusion. This sub-population of dog patients is large, because the incidence rate of nuclear IVD extrusion is very high, especially within specific sub-populations.

SCI can occur at any site in dogs, but is most common adjacent to the thoracolumbar junction (TLJ), probably because of the changing nature of the mechanics and stabilizing influences within this region (19). In dogs, the TLJ contains the spinal cord (which extends into L5 vertebra) and therefore injuries at this level model well the effect of spinal cord lesions in the thoracic part of the human vertebral column. Effects of the lesion on brain control of the distal extremities, including bladder and limbs, can therefore be examined, with the exception of the control of sexual function (which would be difficult to study in dogs).

The shorter general lifespan of dogs implies that canine veterinary patients will become available for post mortem studies much sooner after an intervention than their human counterparts. Nevertheless, dog SCI patients will usually survive many years after their initial injury, providing an opportunity to study the effects of an intervention over a longer follow-up period than is possible in rodents. On the other hand, dogs that have become paraplegic following SCI are susceptible to the same secondary problems as human paraplegics, such as ascending urinary tract infections, providing another opportunity for assessment of the impact of interventions in prehuman clinical trials.

Routine methods of clinical analysis and classification of function in canine SCI patients are very similar to those available for human patients, relying on simple, long-established, scoring systems. However, domestic dogs are highly amenable to sophisticated analysis of movement and the quadrupedal gait can be exploited to provide detailed quantifiable data on the coordination of limb movements between thoracic and pelvic limb girdles. It is particularly helpful in this respect that the most common site of injury in dogs is between T3 and L3 spinal cord segments (and by far most commonly between T12 and L2 (dog spinal cord has 13 thoracic and 7 lumbar segments)) and so lesions at that site disrupt communication between thoracic and pelvic girdles. The quadrupedal gait permits assessment of this coordination in a way that would not be feasible in human subjects and allows precise quantification of deficits and the response to therapeutic interventions (see below).

Any of the feasible approaches for drug delivery in human patients can be used in dogs, including intravenous, intrathecal or even intraparenchymal routes of delivery of both cells and drugs. There are regulations in most countries regarding the ethics of interventions in domestic dogs, but these are not so restrictive as to prevent clinical trials of novel therapies, similar to those that would be permitted in human patients, so long as there is adequately documented scientific justification for the intervention. Regulation of interventions in veterinary patients varies throughout the world and frequently differs from the regulation of experimental procedures carried out in laboratory animals. One obstacle in the UK is the specific prohibition of placebos in clinical trials. This can make it difficult to properly evaluate in a double-blinded study (i.e., when both veterinary investigator and owner are blinded) the implications and treatment efficacy of a novel intervention versus sham therapy. However, for severely injured animals there is (of course) no currently effective therapy, meaning that it is feasible to choose to compare two different plausible interventions based on data obtained in laboratory rodents. Furthermore, there are frequently patient groups that can be identified and act as reasonable, albeit imperfect, controls (for instance paraplegic dogs whose owners choose not to take part in a clinical treatment trial). The specific prohibition of placebos places obstacles on surgical interventions in particular, since it would be very difficult for an Ethical Committee to accept sham surgery for a paralyzed dog.

The severity of functional loss following SCI in dogs has been subject to clinical scoring since the 1950s, using Tarlov scales, or that implemented by Hoerlein (43). Recognizing the need for more precise tools, Olby and others (44) developed a scoring scheme similar to the BBB score used in rodents (7) for use in dogs. Although providing considerably more subtlety, the scoring scheme is not linear, which limits accurate and flexible statistical analysis over the range of possible scores. Furthermore, subjective scores remain susceptible to bias and are often incomplete because they do not take into account urinary and fecal incontinence; also, assessment of fine sensation, which is a component of similar schemes in human patients, is difficult in dogs.

Advanced imaging techniques of the spinal cord such as high field MRI and 16-slice computed tomography (CT) are increasingly available for veterinary patients (45), although dogs must be anesthetized during such examinations. These techniques allow detailed analysis of spinal cord integrity similar to that available in routine clinical practice in humans and can be used to recognize post-traumatic complications of SCI such as syringomyelia (46). Extended regions of hyperintensity on T2-weighted MRI scans are now known to be correlated with poor recovery (45, 47). Diffusion tensor imaging and other less routine MRI techniques, although not commonly used in dogs currently, could also be applied with access to the appropriate software (23).

Electrodiagnostic equipment used for dogs is identical to that used for human patients and similar recordings as can be made in human patients can be obtained from dog patients, providing information on integrity of impulse transmission through the damaged region of spinal cord (48–50). This includes:

There is extensive veterinary experience with these techniques and somatosensory evoked potentials and spinal-evoked potentials have been shown to be clinically relevant to grade SCI in dogs (49, 52). Transcranial magnetic motor evoked potentials (TMMEPs) can be difficult to record in the acute phase of SCI in non-ambulatory dogs (53), but recent investigations have shown that the severity of conduction deficits following TMMEP in dogs with cervical spondylomyelopathy (a chronic compressive lesion of the cervical spinal cord) correlate with the severity of the clinical signs (54). Somatosensory evoked potentials have not been recorded in dogs with thoracolumbar disk herniation and loss of pain sensation caudal to the lesion, but measurements have been made in the acute phase of SCI (49). Recently, we have recorded SSEPs in the chronic phase of SCI in patients that have absent or equivocal pain perception in the hindquarters (41).

Similar to human patients with SCI (55), dogs with SCI will present neurogenic bladder dysfunction such as reflex voiding and detrusor over-reactivity complicated by vesico-sphincterial dysynergia, or detrusor atony (41). Urodynamic equipment used for analysis of human continence can be used in dogs (56) to obtain cystometric, urethral profilometric, and sphincter electromyogram data. This information is directly relevant to understanding possible effects of novel interventions in human patients—for instance, routine urodynamic assessments of function in human SCI patients (57, 58), such as bladder compliance, number of involuntary detrusor contractions, sphincter activity during bladder filling can readily be obtained to define the same aspects of continence in dogs with SCI.

Disruption of quadrupedal coordination after SCI provides an enticing opportunity for functional analysis in dogs. Following severe thoracolumbar SCI, there is complete loss of pelvic limb movement, but often “spinal walking”—defined as stepping by the pelvic limbs in the absence of clinically detectable spinal cord transmission across the lesion—can appear within a few months in dogs (59). A characteristic feature of “spinal walking” is the lack of coordination between thoracic and pelvic limb movements, and there is also a lack of pelvic limb lateral stability. Computerized gait analysis provides methods to quantify both these aspects of spinal walking, meaning that it is possible to attribute a numerical score to the changes in coordination that occur during either spontaneous or intervention-induced recovery (Fig. 4) (for a review see (41)). Using these methods we have shown that normal dogs have extremely rigid coordination between thoracic and pelvic limb motion, and also that this pattern is recovered following mild SCI (41, 60, 61). We are currently using these techniques to evaluate the outcome after intraspinal transplantation of olfactory ensheathing cells (OECs).

Stay updated, free articles. Join our Telegram channel

Join