Neurological Assessment Scores in Rabbit Embolic Stroke Models

INTRODUCTION

Appropriate animal models play critical roles in the development of new treatment strategies for acute stroke, and many have proven successful [1]. Zivin’srabbit model of ischemic stroke in 1985 was crucial to development of successful human tPA stroke therapy [2,3]. Behavioral scores based on assessment of the deficits induced by stroke in animal models are reported in the majority of studies. These scores range from simple (1 or 0 score for deficits present or absent) to more complex (41 point scale). In the most commonly used stroke model, occlusion of the middle cerebral artery of the rat, either transient or permanent, the most frequent battery of behavioral tests had scales of 5, 12, or 18 points [4-6]. These most frequently include paw placing, reflex changes, posture maintenance, circling, paw extension in response to being pushed toward the body, and spontaneous walking. In some scales death is given the maximum score. Similar neurological tests are used to evaluate stroke models in mice, rabbits, and non-human primates. In the rabbit,behavioral scores range from the binary score of 0 or 1 [2] to the more detailed score 0-11 of Zhao [7]. More detailed scores were used inthe photochemical occlusion of the middle cerebral artery method of producing a stroke [7,8]. In those two studies significant differences were found between behavioral scores after experimental treatments compared to non-treated controls. In the present report three previously published methods of producing an embolic stroke in rabbits [9], and the system used to score those animal behaviors for moderate to severe stroke is further evaluated. This review of that scoring system in three different types of embolic stroked non-lytic rabbit groups is necessary to provide proof of the scoring concept for each system’s stroke severity and its accompanying neurological deficits. Resulting deficits were evaluated by a scale similar to the 10 point scale of Kawano et al. [8] in his photochemically induced stroke model and herein called the NAS scale. The NAS scale (or the Wryneck evaluation) has the ability to detect neurological impairment without the prior stimulus training often necessary in rodent models. Our rabbit embolic stroke models using soluble and insoluble emboli produce distinct focal patterns of predominately middle cerebral artery (MCA) ischemic injury. These clinically relevant models mimic situations of soluble and insoluble emboli in humans [9]. We deliver fresh clot [10], aged clot [11], or microsphere emboli [12,13], angiographically into the internal carotid artery (ICA) to cause acute ischemic strokes of moderate severity. Since deaths and inappropriate arterial occlusions are low (≤14%) [9], a comprehensive neurological outcome scale to assess severityof injury was chosen [14]. We investigated whether defined occlusions, vital stains, and histopathologic volume measurements of infarct at 24 hours correlated with neurological outcome assessments.

MATERIALS AND METHODOLOGY

Cerebral Embolic Model

Angiographic procedures were described previously [9-5]. Briefly, rabbits were anesthetized, had angiography and embolization accomplished by injecting soluble clot emboli (fresh, 1.6x 0.6mm, or aged, 4.0 x 1.2 mm) cylindrical clot, or insoluble emboli (700-900 µm microspheres) with 0.7 to 2.0 mL of saline. Repeat angiography 1 minute later documented occlusion (Fig. 1). Neurological deficits were assessed 24 hours later.

RESULTS

NAS Values for Stroke Models

At 24 hr post-embolism mean NAS values were 1.7±0.3 (fresh clot), 2.4±0.5 (aged clot), and 3.6±0.5 (microsphere emboli) and were compared (ANOVA p=0.007). Post hoc, the fresh clot model was different from themicrosphere model (p=0.0018) but neither fresh clot nor microsphere groups were different from the aged clot model. Infarct percent volumes were 0.6±0.2%, 3.3±0.8%, and 5.3±1.2, respectively. A comparison of all NAS values (from each of the three stroke models) vs. %IV values indicates a strong positive relationship (R=0.61, p<0.0001). NAS values resulting from strokes produced in the three individual models correlated with %IV values as follows: fresh clot R=0.501 (Fig. 2), aged clot R=0.654 (Fig. 3), and for microspheres R=0.615 (Fig. 4). Correlations of individual test components to %IV are listed in Table 1.

DISCUSSION

At 24hr after stroke the neurological assessment score (NAS) for all three embolic models significantly correlated with the percent infarct volume(R=0.61, p<0.0001), thus, this behavior assessment can be used as an indicator of the magnitude of neurologic deficits for rabbit embolic stroke models (Table 1). End points in animal models of stroke have included infarct volumes, imaging, blood tests, neurologic deficits, pathological examination, behavior changes and various others, individually or in combination. Almost 15 years ago extensive behavioral evaluations including preconditioning behaviors were recommended for rodents and gerbils along with longer delays before final testing [16-18]. However, even with these evaluations, translation of successful models in small animals to human use often fails [19]. Larger animals, especially rabbits [10] have provided some success, even with very simple behavioral evaluations, perhaps due to issues of brain size, metabolic rate, and ratio of white to gray matter [20].

In our studies, comparison of these three embolic stroke models (fresh clot, aged clot and microsphere) shows corresponding and parallel increases in NAS values and %IV values (Table 2 and Figs. 2-4). Recanalization may be a key factor here. Autolysis of soft fresh clot causing restoration of blood flow is common in our rabbit embolic stroke model while firmer aged clots are resistant to autolysis, and embolic microspheres cause permanent occlusions with no possible autolysis. The fresh clot model usually demonstrates silent anterior infarcts and sparing of the posterior circulation, thereby, causing lower NAS values and lower correlation with the %IV.

In addition, %IV increases with the number of major cerebral vessels occluded and decreases with less effective embolic agents, with corresponding changes in the NAS values in each case. With more vessel occlusions the effective collateral blood flow supporting the penumbra decreases. This may account for larger infarcts and higher NAS and %IV values. The flow directed nature of the angiographic model usually leads to occlusion of the MCA first, the ACA second, with posterior vessels (the posterior cerebral and superior cerebellar arteries) occluding later if more emboli or fragments are involved. Occlusion of the anterior vessels is relatively silent with more dramatic clinical signs associated with the posterior circulation [14]. The NAS confirms this pattern.

Thus, the NAS is a step towards a more complete behavioral characterization for rabbit models of stroke and has now been validated in our three embolic stroke models and by similar assessments in several other different rabbit stroke models. Recent reports confirm the important correlation of acute infarct volume and eventual clinical outcomes in human stroke [20, 21]. This reinforces the value of infarct volumes in all stroke assessments. The strong correlation of overall NAS values with %IV in this study validates the use of the NAS examination in rabbit stroke models, providing an additional indicator of severity without sacrificing the animal or the additive expense of MRI examinations.

CONCLUSION

In summary, the NAS method provides a valid neurologic assessment of ischemic brain infarcts in rabbits. The strong correlation of overall NAS values and the results with %IV in each of our three embolic stroke models here validates the use of NAS in rabbit stroke models without pre-conditioning, sacrificing the animal, or the expense of MRI examinations of the brain.

CONFLICT OF INTEREST:

None

REFERENCES