Abstract
Objectives
Cerebrovascular accidents (CVAs) are infrequently reported in cats. To date, clinical characteristics, including lesion localisation and MRI findings, have only been reported in two cats. The aim of the current study is to document MRI findings in cats presenting with CVAs over an 11 year period. Cases were reviewed according to initial clinical presentation, subsequent physical and neurological findings, predisposing systemic disease and short- and long-term (when available) outcome with a view to identifying any typical pattern in disease occurrence.
Methods
Patient records of cats presenting to a single referral centre from January 2005 to September 2016 with acute onset, non-progressive (after 24 h) intracranial signs compatible with a CVA and where an MRI was performed within 72 h were retrospectively reviewed.
Results
Nine cats met the inclusion criteria. All cats had ischaemic CVAs (presumptively diagnosed in eight cats and confirmed in one cat following post-mortem examination). No cases of haemorrhagic CVAs were identified. Four cats presented with territorial infarcts that were confined to the territory of the rostral or caudal cerebellar arteries (n = 4). Lacunar infarcts were identified in five cats in the location of the cerebrum (n = 1), the thalamus/midbrain (n = 2) and the medulla oblongata (n = 2). Concurrent systemic disease was identified in most (n = 8/9). In the present study short-term prognosis was favourable and 8/9 cats survived to 48 h following admission.
Conclusions and relevance
CVAs in cats occur in the same vascular territories as in dogs and have similar MRI features. This study notes that the presenting cats had a high likelihood of concurrent disease (8/9 cases) but had a favourable short-term prognosis, if neither the clinical presentation nor concurrent disease were severe.
Introduction
Cerebrovascular accidents (CVAs) are defined as a disruption in cerebral blood supply leading to neurological dysfunction and are classically associated with clinical signs that are peracute in onset and non-progressive after 24 h. 1 CVAs, a temporary or permanent reduction in cerebral blood flow, are the most common cause of cerebrovascular disease in both the human population and in veterinary species, and occur either secondary to a local thrombus or a thromboembolism in the case of ischaemic CVAs, or to intracranial blood vessel rupture in haemorrhagic CVAs.1–4 Whereas cerebrovascular disease is the subject of multiple reviews in canine patients,5–8 it has rarely been reported in the cat.9,10 In fact, signalment, concurrent systemic disease and outcome of ischaemic and haemorrhagic CVAs have all been subject to review in dogs.5,6,11,12 To date, clinical characteristics, including MRI findings of ischaemic CVAs, have been reported in just two cats ante-mortem. 9 Cats with cerebrovascular disease have also been described in relation to histopathological features, including seven cases of ischaemic CVAs (six diagnosed following post-mortem examination and a further one by CT-guided biopsy) and five haemorrhagic CVAs. 10 Post-mortem reports introduce inherent bias in selecting the most severely affected cases. Therefore, it is unclear whether these previous reports represent an accurate reflection of CVA distribution and frequency in cats.
The MRI characteristics of CVAs in dogs are well described, 5 whereas in cats they are limited to a single report of two cases of rostral cerebellar infarction. 9 Species-specific adaptations in cerebral blood supply are well described; for example, it is known that the major blood vessels supplying the arterial circle in the canine and feline populations differ.3,10,13–15 Compared with cats, the vertebral arteries supply a much greater proportion of the canine brain, including the thalamus and caudal-most aspect of the cerebral cortex.15–17 In cats the vertebral artery irrigates only the caudal-most part of the medulla oblongata. 14 Whether these anatomical differences influence the frequency or distribution of infarcts in cats compared with dogs, or alter typical MRI features is currently unclear.
The aim of the current study is to report the MRI features, clinical characteristics (inclusive of concurrent systemic disease), and short- and long-term outcome of CVAs in the feline population presenting to a single referral hospital over an 11 year period.
Materials and methods
Approval for the study was obtained from the Clinical Research Ethical Review Board at the Royal Veterinary College. Medical records of all cats presenting to a single referral hospital (Queen Mother Hospital for Animals, Royal Veterinary College) between January 2005 and September 2016 with acute onset, non-progressive (after 24 h) intracranial signs were retrospectively reviewed for inclusion in the study. From these, only cats that presented within 48 h of onset of clinical signs, had full physical and neurological examinations and underwent MRI within 72 h of presentation were included. Clinical records were systematically reviewed to take into account the following factors: (1) presentation; (2) clinical progression; and (3) additional diagnostic tests used to assess for concurrent disease. Where available, cerebrospinal fluid (CSF) analysis (cell count, cytological examination, total protein concentration determination and infectious disease testing) was recorded. Of those with CSF analysis available, only cats that had results consistent with previous reports of a CVA, specifically normal CSF analysis or mild neutrophilic or mononuclear pleocytosis (<30 cells/μl), 6 were included.
All MRI sequences were obtained with a 1.5 Tesla unit (Intera Pulsar System; Philips Medical Systems). A complete MRI series included the following sequences: T2- weighted images (T2W) in the sagittal and transverse planes and pre- and post-contrast T1-weighted images (T1W) in transverse planes. Other sequences were analysed when available and included T2W and T1W sequences in the dorsal plane, fluid-attenuated inversion recovery (FLAIR), gradient echo (GRE) and diffusion-weighted imaging (DWI). Cats that received corticosteroids or antibiotics, other than those that received a single dose, were excluded from the study. The post-contrast enhancement was categorised as none, mild when contrast enhancement was hypointense compared with the fat tissue on T1W images and marked when the contrast enhancement was isointense to the fat tissue on T1W images. The contrast enhancement was also categorised as homogeneous when there was uniform enhancement and heterogeneous when it was dissimilar throughout the enhanced area. The degree of contrast enhancement in lesions was assessed and classified as none, mild or marked and heterogeneous or homogeneous by two independent reviewers.
In all cases, a board-certified veterinary radiologist (RD) and a board-certified veterinary neurologist (EB) independently reviewed the MRI sequences available. A diagnosis of CVA was made when both reviewers independently agreed on a diagnosis, according to imaging criteria outlined previously. 5 In the case of ischaemic CVAs, this included lesions that were predominantly confined to the grey matter and limited to the vascular territory of a main artery (territorial CVA) or perforating arteries (lacunar CVA). Lesions were predominantly homogeneous in appearance and hyperintense on T2W and FLAIR sequences with well-demarcated boundaries to adjacent brain parenchyma. Minimal or no mass effect was noted. Cats presenting with suspected CVAs affecting thalamic arterial blood supply were further categorised in accordance with a classification scheme previously applied in dogs that subdivides thalamic infarcts by location and vascular territory into: (1) paramedian, (2) dorsal extensive and (3) ventrolateral CVAs. 4 In the case of haemorrhagic CVAs, MRI sequences were reviewed to identify the presence of acute or early subacute intraparenchymal haemorrhage identified by iso- to hyperintense signal on T2W images, and iso- to slightly hyperintense signal on T1W images. The presence or absence of mass effect was recorded. GRE sequences were reviewed when available and considered compatible with haemorrhage when lesions were hypointense. 18 Cats with other brain pathology were excluded.
Reviewers (EB, RD) further defined infarcts according to location (cerebrum, thalamus, midbrain, cerebellum [divided into rostral and caudal cerebellum according to their location in relation to the primary fissure], pons, medulla oblongata) and arterial territory in the cases of ischaemic CVAs classified as territorial (occlusion of a main artery, including rostral cerebral, middle cerebral, caudal cerebral, rostral cerebellar, caudal cerebellar or vertebral) or lacunar (occlusion of a small perforating arterial vessel, including, superficial and deep perforating arteries).
Short-term outcome was defined as survival or non-survival at 48 h after presumptive diagnosis of a CVA. Long-term outcome was assessed, when possible, at a minimum of 4 weeks after diagnosis. Follow-up information was obtained by telephone consultation with the owner. Long-term outcome was defined as dead or euthanased, poor (severe neurological deficits or no improvement in neurological signs since presentation), good (clinical signs improved since discharge but neurological deficits still present) or excellent (clinical signs resolved since discharge and the cat considered normal).
Results
Fourteen cats presented to the Queen Mother Hospital for Animals, between January 2005 and September 2016 with suspected CVAs. Two cats were excluded as their clinical signs improved and they were discharged without MRI. A further three cats were excluded, two owing to the additional finding of a meningioma and a further was excluded having received a prolonged course of corticosteroids on the basis that inflammatory, infectious and neoplastic disorders could not be excluded in this case. A total of nine cats were evaluated for this study. Haemorrhagic CVAs were not identified in any case. In the remainder of the text the term ‘CVA’ in this study should be construed as referring only to ischaemic CVAs (presumptively diagnosed in eight cats and confirmed in one following post-mortem examination). All eight cases with suspected CVAs displayed improvements in clinical signs prior to discharge. In the confirmed case, the owner elected for euthanasia owing to the severity of clinical signs following cardiorespiratory arrest despite successful resuscitation. In this case a cerebellar ischaemic CVA was confirmed post mortem.
Of those cats included there were four (44.5%) neutered females, one (11.1%) sexually intact female and four (44.5%) neutered males with a median age of presentation of 12 years (median 12 years, range 2 years–16 years). Breeds represented included domestic shorthair (n = 7), domestic longhair (n = 1) and Bengal (n = 1).
Imaging findings
All cats presented within 48 h and were imaged within 72 h of onset of clinical signs. T2W sequences in the sagittal and transverse planes and pre- and post-contrast T1W images in the transverse plane were available in all cats. Additional sequences were also obtained, including T2W dorsal (n = 2), FLAIR (n = 6), GRE (n = 8) and DWI (n = 4). All lesions were focal and well to marginally well demarcated to adjacent brain parenchyma. Lesions were present in the location of a vascular territory and predominantly restricted to grey matter in all but one case (Figures 1 and 2). In that case the cerebral lesion extended from the head of the caudate nucleus to the white matter of the internal capsule (Figures 1a,c and 2b,c). In all cases, lesions were hyperintense to normal grey matter on T2W and, when available, on FLAIR sequences and were either slightly heterogeneous in the periphery (n = 1) or homogeneous (n = 8) in appearance. In T1W sequences lesions were isointense (n = 3), hypointense (n = 4) or hypo-isointense (n = 2) to normal grey matter. Lesions either exhibited no contrast enhancement (n = 3), mild to no contrast enhancement (n = 4) or mild enhancement that was restricted to the periphery of the lesion (n = 2). Lesions were hyperintense to normal grey matter on GRE (n = 5) with no signal void. Where DWI was available (n = 4), lesions were hyperintense to normal grey matter on DWI and hypointense on the apparent diffusion coefficient map, indicative of restricted diffusion, consistent with cytotoxic oedema (Figure 3). 19 DWI was non-diagnostic in one case. There was no (n = 5) or minimal (n = 4) mass effect in all cases.
Figure 1.
Feline cerebrovascular accidents were identified in multiple brain regions (a–c). (a) Schematic representation of the ventral surface of a cat brain and cranial cervical spinal cord showing the cerebral blood supply, rostral to the top. The distribution of territorial and lacunar infarcts in the cerebral arteries and their branches identified in clinical cases is shown (1–5) where blue circles indicate the approximate location of suspected infarcts. (b,c) Schematic representation of the midline sagittal view (b) and transverse view at the level of the caudate nucleus, of a cat brain shown in (a). Note the distribution of lesions corresponding to the labels in (a)
Figure 2.
MRI (sagittal and transverse T2-weighted images [T2W]) of the brain in cats with presumed cerebrovascular accidents (a–m). Note that the position of transverse slices is indicated in (a). Lesions are focal, well-demarcated T2-weighted hyperintense. Examples are shown of those in the internal capsule (b,c), the paramedian thalamus (d,e), the midbrain (f,g), the rostral and middle cerebellum (h,i), the caudal cerebellum (j,k) and the caudal medulla oblongata (l,m)
Figure 3.
MRI of the brain of a 13-year-old female neutered domestic shorthair cat. T2-weighted (T2W) sagittal (a) and transverse images (b), diffusion-weighted imaging (DWI) (c) and a corresponding apparent diffusion coefficient (ADC) map (d) of a presumed cerebrovascular accident in the rostral cerebellum. The lesion was focal, well-demarcated, hyperintense to normal grey matter on T2W images, hyperintense on DWI and hypointense on ADC map indicative of restricted diffusion
The distribution of suspected CVAs was determined based on MRI findings and as follows: cerebral (extending into the thalamus, n = 1/9; thalamic/midbrain, n = 2/9; cerebellar, n = 4/9; and those restricted to the medulla oblongata, n = 2/9 [Table 1]). Lesions were restricted to the following vascular territories as determined by MRI: lenticulostriate arteries of the middle cerebral artery (n = 1/9), perforating arteries of the caudal cerebral artery (n = 2/9), rostral cerebellar artery (n = 2/9), caudal cerebellar artery (n = 1/9), rostral and caudal cerebellar arteries (n = 1/9) and basilar artery (n = 2/9) (Figure 1).
Table 1.
Neurological signs in cats according to vascular territory
| Vascular territory (n) | Clinical signs (n) |
|---|---|
| Striate artery (of the middle cerebral artery) (1) | Altered mentation (obtunded) (1), anisocoria (1), circling (1), postural reaction deficits (1) |
| Perforating artery of the caudal cerebral artery (2) | Altered mentation (obtunded) (2), head tilt (1), non-ambulatory (1), anisocoria with altered pupillary light reflex (1), reduced menace (1), decreased vestibulo-ocular reflex (1), decreased postural reactions (2), circling (1), cervical hyperaesthesia (1) |
| Rostral cerebellar artery (1), caudal cerebellar artery (1), multifocal (rostral and caudal cerebellar arteries) (2) | Decerebellate posture (2), reduced–absent menace with intact vision (2), head tilt (2), nystagmus (3), vestibular ataxia (2), cerebellar ataxia (1), intention tremor (1), reduced postural reactions (1) |
| Perforating branches of the basilar artery (2) | Altered mentation (obtunded) (2), proprioceptive ataxia (1), ambulatory tetraparesis (1), decreased menace (2), cranial nerve deficits (reduced facial sensation, reduced palpebral) (1), anisocoria (1), cervical hyperaesthesia (1). |
The cerebral lesion was identified at the level of the head of the caudate nucleus and extended into the adjacent internal capsule (Figure 2b,c) in the arterial territory of the lenticulostriate arteries (a branch of the middle cerebral artery) (Figure 1a,c). The thalamic and midbrain lesions were identified in the right ventral thalamus at the level of the pituitary gland and in the left midbrain at the level of the rostral colliculi (Figures 1b, 2d–g). These presumptive CVAs were located in the territory of perforating arteries (branches of the caudal cerebral artery), namely the proximal and caudal perforating arteries, respectively (Figure 1a,b). Those confined to the proximal and caudal perforating arteries affecting the thalamic arterial blood supply were further defined topographically as paramedian, according to the classification scheme applied to dogs. 7
Lesions were identified in the cerebellum of four cats (Figure 1b). In two cats a lesion was identified in the territories of the right and left rostral cerebellar arteries affecting the rostral and middle cerebellum (Figure 2h,i), whereas another cat had a lesion in the right caudal cerebellum in the territory of the right caudal cerebellar artery (Figure 2j,k). One cat had a lesion in the vascular territory of both the rostral and caudal cerebellar arteries (Figure 1a,b). A further two cats had CVAs in the medulla oblongata in the territory of the perforating branch of the basilar artery (Figures 1a,b and 2l,m). One cat had a lesion in a rostrolateral position in the upper rostral medulla oblongata, whereas a further cat had a lesion in the dorsomedial location in the middle medulla oblongata.
Clinical signs
In all cases, clinical signs reflected neurological dysfunction and varied dependent on the location of the CVA (summarised in Table 1). Specifically, the cat presenting with a presumptive diagnosis of a CVA in the middle cerebral artery and those cats presenting with a presumed CVA in the caudal cerebral artery exhibited signs of forebrain dysfunction, those located in the rostral and caudal cerebellar arteries exhibited signs of cerebellar dysfunction and those with a presumed CVA located in the basilar artery exhibited signs of brainstem dysfunction.
General physical examination was unremarkable in all but three. Two cats had a grade II/VI heart murmur and one had a gallop rhythm noted on thoracic auscultation. Those two cats with a heart murmur were diagnosed with systemic hypertension with bilateral retinal haemorrhage noted on fundoscopy. Overall, concurrent disease was identified in 8/9 (89%) cats (Table 2). Of these, four cats (44%) had hypertrophic cardiomyopathy (HCM). Of those with HCM two had normal left atrial size, one had a dilated left atrium with severe left ventricular hypertrophy and ventricular outflow tract obstruction, the other had marked dilatation of the left atrium with poor left atrial formation and a thrombus was also identified. One of the two cats with HCM and normal left atrial size had concurrent systemic hypertension as outlined above. Of the five cats without a diagnosis of HCM, two had HCM ruled out by echocardiogram, although left ventricular hypertrophy was noted on post-mortem examination in one. Two did not undergo diagnostic cardiac ultrasound or post-mortem examination (HCM diagnoses are therefore unknown, one cat in this group had a grade II/VI heart murmur as mentioned previously).
Table 2.
Concurrent medical conditions in cats according to vascular territory
| Vascular territory (n) | Concurrent medical conditions (n) |
|---|---|
| Striate artery (middle cerebral) (1) | None (1) |
| Perforating artery (caudal cerebral) (2) | Pulmonary mass with metastases (1), hypertension (1) |
| Rostral cerebellar artery (1) | Pulmonary adenocarcinoma (1) |
| Caudal cerebellar artery (1) | Hypertrophic cardiomyopathy (left atrial enlargement and thrombus present) (1) |
| Rostral and caudal cerebellar arteries (2) | Hypertrophic cardiomyopathy (with no sign of left atrial enlargement) (1), hyperthyroidism (previously diagnosed and receiving treatment) and bilateral otitis media (1) |
| Vertebral artery (2) | Hypertrophic cardiomyopathy (with left atrial enlargement) (1), hypertrophic cardiomyopathy (no sign of left atrial enlargement) with hypertension (1) |
Of those cats that were not diagnosed with HCM, two had pulmonary masses. One cat was diagnosed with a mass in the right caudal lung lobe that was suspected to have metastasised to a digit. Subsequent cytological examination of the suspected metastatic mass was consistent with a carcinoma suspected to have metastasised from a primary pulmonary carcinoma. The cat was euthanased owing to a poor prognosis; however, post-mortem examination was not performed. A pulmonary adenocarcinoma was also identified in one cat following post-mortem examination (euthanased following cardiopulmonary arrest after a rostral cerebellar CVA). There were no reported clinical signs attributable to the mass on presentation.
One cat had been diagnosed with hyperthyroidism 1 year prior to presentation to the referral hospital and was on carbimazole. The thyroid status of the cat at the time of presentation was unknown. In that case, bilateral otitis media was also detected by MRI.
CSF findings
CSF was available in 4/9 of cases. All cats had results consistent with previous reports of CVAs and were therefore eligible for inclusion in the study. Of these, CSF was reported as normal in 3/4 (total nucleated cell count [TNCC] <5 mm3, protein <0.25 g/l). An increased TNCC count was identified in one cat (TNCC 18 mm3, red blood cells 0 mm3, protein 0.01 g/l). In that case Toxoplasma gondii was negative by PCR in CSF. A differential count in this case revealed three nucleated cells, specifically two large mononuclear cells and one macrophage with no sign of an aetiological agent.
Outcome
Most cats (8/9; 89%) had a good short-term prognosis (survived to 48 h). One cat with a suspected rostral cerebellar CVA was euthanased and the presumptive diagnosis of a cerebellar CVA according to MRI findings was subsequently confirmed by post-mortem examination. The cat had initially suffered a number of cardiorespiratory arrests following anaesthesia and, although each time was successfully resuscitated, the cat was subsequently euthanased. Post-mortem examination in this case revealed moderate left ventricular hypertrophy with evidence of renal infarcts and a pulmonary adenocarcinoma with angioinvasion. One cat survived to 48 h, and was discharged following an improvement in clinical signs but was re-admitted 5 days after initial discharge owing to inappetence. Further investigation revealed a suspected primary pulmonary mass with likely secondary metastasis to a digit. Cytological analysis of the metastatic mass was consistent with a carcinoma with evidence of necrosis and osteolysis. This cat was euthanased owing to poor prognosis. Long-term follow-up information was available for 2/7 (29%) remaining cases ranging from 5–10 weeks. One cat was deemed to have an excellent outcome following discharge according to the owners (complete resolution of clinical signs) and one cat had a good long-term outcome as there was a reported improvement in clinical signs; however, some neurological deficits were still present (mild head tilt and mild vestibular ataxia).
Discussion
This study provides the most comprehensive overview to date of MRI and clinical characteristics of CVAs in cats. It is the first ante-mortem study to review the MRI characteristics of naturally occurring CVAs in the feline population other than CVAs occurring in the cerebellum. In agreement with a previous post-mortem study, 10 these data demonstrate that feline CVAs can occur in a number of vascular territories.
The retrospective nature of the study introduces significant limitations. In a recent study on dogs with histologically confirmed intracranial disease, MRI was 97.7% specific for identification of cerebrovascular lesions; 20 however, inter-observer agreement in CVAs was only fair and, as such, strict inclusion criteria were applied to this dataset, that is, improvement of clinical signs in the absence of specific treatment in all cases where post-mortem examination was unavailable. In the present study a full MRI sequence was available in all cases (T2W in the sagittal and transverse planes and pre- and post-contrast T1W in transverse planes). Although additional MRI sequences provide further support of a diagnosis of CVA, such as DWI, these sequences were only obtained in 4/9 cats in this study. In prospective studies the inclusion of additional MRI sequences such as perfusion-weighted imaging – a sequence that compliments DWI by permitting analysis of the blood supply of the tissue21,22 – may also be useful given that post-mortem confirmation is rarely available.
It is understood that the incidence of CVAs in cats and dogs is significantly lower when compared with the high incidence of strokes in the human population (approximately 795,000 individuals affected in the USA per year). 23 It has previously been proposed that the lower incidence of CVAs in dogs compared with humans is, in part, related to differences in cerebral blood supply. Specifically, dogs have a higher number of intra-extracranial arterial anastomoses, 16 which reflects an anatomical variation that is also recognised in cats. 24 Thus, species-specific adaptations of the cerebral blood supply present a credible explanation for differences in prevalence and severity of CVAs. Here, nine cats with suspected CVAs were identified in a single referral centre over an 11 year period (compared with 40 cases in dogs in a single report over a 52 month period). 5 Given the lack of data on the total population of cats and dogs presenting to the respective referral centres, it is unclear if this comparison infers a higher incidence of CVAs in cats vs dogs. However, assuming the data represent some evidence of differentiation between the species with respect to the occurrence of CVAs, the literature has to date failed to offer an explanation. As stated, in common with dogs, cats have a high number of intra-extracranial anastomoses,13,14 but despite this commonality, the cerebral blood supply differs between the dog and the cat. Specifically, the main arterial blood supply to the brain differs; the internal carotid and basilar arteries are the major contributors to the cerebral arterial circle in dogs, 15 whereas the maxillary artery (derived from the external carotid artery) supplies all but the caudal region of the medulla oblongata (to the caudal boundary of the trapezoid body) in cats.13,14,24 It remains unclear whether this difference protects the feline population against CVAs, and, if so, how, or whether the lower incidence can be attributed to a different reason, for example prevalence of predisposing disease. It is possible that fewer cats are referred for further investigation of CVAs; either owing to less severe and rapidly improving or resolving neurological signs or to significant concurrent medical conditions resulting in owners more readily electing euthanasia. Further surveys of first-opinion (primary) practices are required to investigate this possibility.
Whereas supratentorial territorial CVAs are most commonly reported in human patients, with cerebellar CVAs accounting for approximately 1.5% of all strokes only, 25 by contrast and in common with dogs, this study suggests that territorial CVAs in cats are more likely to occur in the cerebellar arteries. 5 However, owing to the small sample size in this study caution should be taken in overstating this comparison. Four cats presented with territorial CVAs, all of which occurred in the cerebellum, whereas five cats had lacunar CVAs. A previous ante-mortem study also identified rostral cerebellar CVAs in two cats. 9 Frequency of territorial CVAs in the aforementioned ante-mortem report and in this study are in contrast to an earlier post-mortem study, which did not identify a single CVA in the cerebellum. 10 That study reviewed seven cats with ischaemic CVAs and instead identified five instances of territorial infarcts in the region of the cerebrum and brainstem in addition to two lacunar infarcts affecting the cerebrum and brainstem. 10 The contrast between the current report and previous post-mortem studies may be owing to inherent differences in selection of cases in ante-mortem vs post-mortem studies with those included in the latter more likely to have severe neurological signs or unconnected and potentially severe underlying disease and/or clinical signs.
Cerebellar infarcts appear to occur predominantly in the territory of the rostral cerebellar artery in dogs. To our knowledge there has only been a single case report describing a suspected caudal cerebellar artery infarct in a dog. 26 The current report identifies an equal incidence of infarcts in both the rostral and caudal cerebellar territories in cats (n = 3 in each case). This difference in distribution cannot be ascribed to a differentiation in the direct blood supply of the cerebellum between the species, given the blood supply of the cerebellum in cats is similar to dogs; the rostral and caudal cerebellar arteries supply the rostral and caudal cerebellum, respectively. However, differences have been noted in the blood supply of the caudal brainstem from which the caudal cerebellar artery is derived. 24 In common with dogs, cats have extensive anastomoses supplying the brainstem derived from the cervical region. 24 In the dog (but not in the cat) the occipital artery provides significant blood supply to the brainstem by contributing to the vertebral artery. 24 By contrast, in cats the basilar artery arises in the caudal medulla derived solely from the vertebral arteries without contribution from the occipital artery. It is plausible, although not proven dispositive, that this anatomical divergence in brainstem blood supply contributes to the potentially more frequent instances of caudal cerebellar CVAs in cats given the degree of collateral blood supply in the dog.
Cerebellar infarcts in human patients are, in general, most commonly linked to cardiac disease. Cerebral embolism is reported as the prevailing cause of an infarct in the superior cerebellar artery (equivalent to rostral cerebellar artery), 25 which is predominantly linked to cardiogenic causes such as valvular disease, dysrhythmias and myocardial infarction, 27 whereas posterior inferior cerebellar infarcts (equivalent to the caudal cerebellar artery in cats) were caused by cardiogenic embolism and posterior circulation arterial disease, including atherosclerotic disease (secondary to occlusion or stenosis of the vertebral or the posterior inferior arteries) 28 in equal proportions.25,29 In this study, HCM was identified in four cats with concurrent systemic hypertension reported in one. Left atrial enlargement, a sequel of left ventricular hypertrophy in HCM, leads to secondary blood stasis and, as such, predisposes cats to thromboembolic disease. Of the four cats diagnosed with HCM included in the present study, left atrial enlargement was identified in two and of these a thrombus was identified in one. That cat, presumptively diagnosed with a CVA in the caudal cerebellar artery, had marked left atrial dilation with poor left atrial function. In the absence of left atrial enlargement, left ventricular hypertrophy is unlikely to lead to thromboembolic disease and therefore HCM is likely an incidental finding in the two remaining cases. One cat with caudal and rostral cerebellar CVAs had mild HCM with no dilation of the left atrium. Of the remaining two cats with presumed cerebellar CVAs, one had a primary pulmonary mass and the other had been diagnosed with hyperthyroidism 1 year earlier and was receiving carbimazole. The thyroid status of this cat was unknown and a full echocardiogram could not be performed, although a limited examination revealed mild septal hypertrophy deemed unlikely to be clinically significant. It remains unclear whether correlation noted in this study between concurrent disease and CVAs is causative and/or a factor that influences the site of infarction in cats or merely coincidental given the median age of the cat population is 12 years.
Two cats presented with presumptive CVAs in the medulla oblongata, which is a rare finding in dogs. 5 In fact, it has only been reported in three dogs with multifocal lesions with presumptive CVAs occurring in both the thalamus and in the medulla oblongata. 5 This is in common with human patients where CVAs in the medulla oblongata are rare. 30 The occurrence in cats appears higher, accounting for 2/9 of cases in this study. The blood supply of the caudal brainstem differs between dogs and cats with the vertebral artery receiving a contribution from the occipital artery in dogs. The degree of collateral blood supply therefore offers a plausible explanation for the apparent low prevalence of medullary CVAs in the dog.
Suspected thalamic CVAs were identified in two cats secondary to a lacunar CVA affecting the proximal perforating artery. These suspected CVAs were further classified as paramedian, the most common thalamic CVAs noted in dogs. 7 In dogs with presumed thalamic CVAs, the most common presenting signs were attributable to vestibular dysfunction; vestibular ataxia was reported most frequently irrespective of the position of the thalamic CVA. Vestibular ataxia was not noted in either cat, although one, with a lesion in the territory of the caudal perforating artery (midbrain), was non-ambulatory and displayed a predominance of vestibular signs, including a head tilt (Table 1). The second cat, with a lesion in the territory of the proximal perforating artery at the level of the interthalamic adhesion, was obtunded. The cat circled ipsilateral to the lesion and had contralateral postural reaction deficits (Table 1). Given the low numbers of cats with presumptive thalamic CVAs it is not possible to determine if the recognised clinical syndromes noted in dogs can also be applied to cats.
In dogs, there was no observable statistical correlation between the site of the CVA and a specific medical condition, although a single study of dogs with ischaemic CVAs revealed that 50% had concurrent systemic disease. 6 The predominant medical conditions in these dogs were chronic kidney disease and untreated hyperadrenocortism, affecting eight (44%) and six (33%) of the 18 dogs, respectively. None of the cats in the present report was diagnosed with renal disease, which is particularly noteworthy given the high general prevalence of renal disease in cats, particularly in the ageing feline population.
It is difficult to draw reliable conclusions regarding disease prevalence between dogs and cats given the small referral cohort included. However, it is worth noting that cats and dogs with presumed CVAs tend to present at an older age (the median age of cats included in this study was 12 years of age, whereas the median age of dogs was 8 years). 5
Pre-existing systemic disease was common in these cats presenting with CVAs. With 8/9 cats presenting with concurrent systemic disease, this appears more common than previously reported in dogs (50%). However, it is noteworthy that in two cats, HCM was mild with no left atrial enlargement and is therefore unlikely to predispose to thromboembolic disease. Two cats with cerebellar CVAs have been previously described ante-mortem. 9 Interestingly, both cats had concurrent medical conditions that may increase the risk of thromboembolic disease, specifically restrictive cardiomyopathy with valvular endocarditis and pulmonary neoplasia. 9 In a separate study conducted post-mortem, extra-cranial pathology was identified in 5/7 cats with ischaemic CVAs, including HCM and lymphoma. 10 Other studies have shown that Cuterebra species larval migration is associated with cerebrovascular disease in cats, although pathology was not restricted to ischaemic CVA alone31,32 Therefore, a cat presenting with signs consistent with a CVA warrants a full systemic investigation. Most cats (8/9) survived to 48 h following diagnosis and therefore had a good short-term outcome. As mentioned, one cat was euthanased owing to the severity of clinical signs. Post-mortem examination in this case revealed an extensive CVA in the territory of the rostral cerebellar artery in addition to moderate HCM, pulmonary adenocarcinoma and evidence of infarcts in the kidneys suggestive of a increased likelihood of thromboembolism. A further cat was euthanased (after an initial improvement in clinical signs) owing to the severity of the concurrent medical condition (pulmonary mass with metastases) despite clinical improvement in neurological signs. These data suggest that short-term prognosis is good if the cat is neither severely affected nor impacted by critical concurrent disease. Previous reports in dogs have shown that prognosis is good if no underlying disease is found. In a single study with a total of 22 dogs with presumed CVAs, 23% died within 30 days; however, those that survived this period had a median survival time of 505 days. 11 The low number of cases and the retrospective nature of the study make it impossible to determine long-term prognosis in cats with CVAs. In the cases observed in this study, follow-up information for long-term outcome was only available in 2/7 cats that survived to discharge; of those two, one had a good and the other an excellent outcome at a minimum of 5 weeks, although information beyond this time period is only sporadically available and conclusions cannot be drawn. This highlights a further limitation of the study owing to its retrospective nature. It should be noted that all the cats included had neurological signs that improved prior to discharge. These inclusion criteria may inadvertently introduce bias in selecting for less severe cases, which may serve to exclude cats with extensive lesions or severe disease requiring more time before significant improvement is identified. While it would be useful to document the long-term prognosis in these cases, the study period reviewed (11 years) and apparent low prevalence of ischaemic CVAs in cats means that further collection of long-term data would likely be problematic. Given the higher occurrence of concurrent disease identified in these cats, it is possible to argue that prognosis for CVAs is more guarded for the feline population in general, although this will be dependent on the severity and type of the relevant concurrent disease. A further study is necessary to determine the long-term outcome of these cases.
Conclusions
Ischaemic CVAs appear to be an uncommon cause of neurological disease in cats. When they do occur, these data suggest they are most commonly located caudal to the tentorium, and, in particular, the cerebellum. All regions of the brain, however, may be affected and therefore vascular disease needs to be considered in any cat presenting with peracute-onset non-progressive (over 24 h) intracranial signs. Concurrent systemic disease in these cats was common and in particular HCM with or without concurrent systemic hypertension. Despite this, the chance of recovery to discharge is favourable assuming neither the clinical presentation nor concurrent disease are severe.
Footnotes
Accepted: 6 July 2017
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The authors received no financial support for the research, authorship, and/or publication of this article.
References
- 1. Hankey GJ. Stroke. Lancet 2016; 389: 641–654. [DOI] [PubMed] [Google Scholar]
- 2. Garosi LS. Cerebrovascular disease in dogs and cats. Vet Clin North Am Small Anim Pract 2010; 40: 65–79. [DOI] [PubMed] [Google Scholar]
- 3. Fankhauser R, Luginbuhl H, McGrath JT. Cerebrovascular disease in various animal species. Ann N Y Acad Sci 1965; 127: 817–60. [DOI] [PubMed] [Google Scholar]
- 4. Wessmann A, Chandler K, Garosi L. Ischaemic and haemorrhagic stroke in the dog. Vet J 2009; 180: 290–303. [DOI] [PubMed] [Google Scholar]
- 5. Garosi L, McConnell JF, Platt SR, et al. Clinical and topographic magnetic resonance characteristics of suspected brain infarction in 40 dogs. J Vet Intern Med 2006; 20: 311–321. [DOI] [PubMed] [Google Scholar]
- 6. Garosi L, McConnell JE, Platt SR, et al. Results of diagnostic investigations and long-term outcome of 33 dogs with brain infarction (2000–2004). J Vet Intern Med 2005; 19: 725–731. [DOI] [PubMed] [Google Scholar]
- 7. Goncalves R, Carrera I, Garosi L, et al. Clinical and topographic magnetic resonance imaging characteristics of suspected thalamic infarcts in 16 dogs. Vet J 2011; 188: 39–43. [DOI] [PubMed] [Google Scholar]
- 8. McConnell JF, Garosi L, Platt SR. Magnetic resonance imaging findings of presumed cerebellar cerebrovascular accident in twelve dogs. Vet Radiol Ultrasound 2005; 46: 1–10. [DOI] [PubMed] [Google Scholar]
- 9. Cherubini GB, Rusbridge C, Singh BP, et al. Rostral cerebellar arterial infarct in two cats. J Feline Med Surg 2007; 9: 246–253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Altay UM, Skerritt GC, Hilbe M, et al. Feline cerebrovascular disease: clinical and histopathologic findings in 16 cats. J Am Anim Hosp Assoc 2011; 47: 89–97. [DOI] [PubMed] [Google Scholar]
- 11. Gredal H, Toft N, Westrup U, et al. Survival and clinical outcome of dogs with ischaemic stroke. Vet J 2013; 196: 408–413. [DOI] [PubMed] [Google Scholar]
- 12. Lowrie M, De Risio L, Dennis R, et al. Concurrent medical conditions and long-term outcome in dogs with non-traumatic intracranial hemorrhage. Vet Radiol Ultrasound 2012; 53: 381–388. [DOI] [PubMed] [Google Scholar]
- 13. Kamijyo Y, Garcia JH. Carotid arterial supply of the feline brain. Applications to the study of regional cerebral ischemia. Stroke 1975; 6: 361–369. [DOI] [PubMed] [Google Scholar]
- 14. Holmes RL, Newman PP, Wolstencroft JH. The distribution of carotid and vertebral blood in the brain of the cat. J Physiol 1958; 140: 236–246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Jewell PA. The anastomoses between internal and external carotid circulations in the dog. J Anat 1952; 86: 83–94. [PMC free article] [PubMed] [Google Scholar]
- 16. Garosi LS, McConnell JF. Ischaemic stroke in dogs and humans: a comparative review. J Small Anim Pract 2005; 46: 521–529. [DOI] [PubMed] [Google Scholar]
- 17. Wellens DL, Wouters LJ, De Reese RJ, et al. The cerebral blood distribution in dogs and cats. An anatomical and functional study. Brain Res 1975; 86: 429–438. [DOI] [PubMed] [Google Scholar]
- 18. Bradley WG., Jr. MR appearance of hemorrhage in the brain. Radiology 1993; 189: 15–26. [DOI] [PubMed] [Google Scholar]
- 19. Moseley ME, Cohen Y, Mintorovitch J, et al. Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Reson Med 1990; 14: 330–346. [DOI] [PubMed] [Google Scholar]
- 20. Wolff CA, Holmes SP, Young BD, et al. Magnetic resonance imaging for the differentiation of neoplastic, inflammatory, and cerebrovascular brain disease in dogs. J Vet Intern Med 2012; 26: 589–597. [DOI] [PubMed] [Google Scholar]
- 21. Moustafa RR, Baron JC. Clinical review: imaging in ischaemic stroke – implications for acute management. Crit Care 2007; 11: 227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Schellinger PD, Bryan RN, Caplan LR, et al. Evidence-based guideline: the role of diffusion and perfusion MRI for the diagnosis of acute ischemic stroke: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 2010; 75: 177–185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Mozaffarian D, Benjamin EJ, et al. Heart disease and stroke statistics – 2016 update: a report from the American Heart Association. Circulation 2016; 133: e38–360. [DOI] [PubMed] [Google Scholar]
- 24. Gillilan LA. Extra- and intra-cranial blood supply to brains of dog and cat. Am J Anat 1976; 146: 237–253. [DOI] [PubMed] [Google Scholar]
- 25. Kase CS, Norrving B, Levine SR, et al. Cerebellar infarction. Clinical and anatomic observations in 66 cases. Stroke 1993; 24: 76–83. [DOI] [PubMed] [Google Scholar]
- 26. Negrin A, Gaitero L, Anor S. Presumptive caudal cerebellar artery infarct in a dog: clinical and MRI findings. J Small Anim Pract 2009; 50: 615–618. [DOI] [PubMed] [Google Scholar]
- 27. Stirling J, Muramatsu K, Shirai T. Cerebral embolism as a cause of stroke and transient ischemic attack. Echocardiography 1996; 13: 513–518. [DOI] [PubMed] [Google Scholar]
- 28. Nouh A, Remke J, Ruland S. Ischemic posterior circulation stroke: a review of anatomy, clinical presentations, diagnosis, and current management. Front Neurol 2014; 5: 30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Amarenco P, Kase CS, Rosengart A, et al. Very small (border zone) cerebellar infarcts. Distribution, causes, mechanisms and clinical features. Brain 1993; 116: 161–186. [DOI] [PubMed] [Google Scholar]
- 30. Kim K, Lee HS, Jung YH, et al. Mechanism of medullary infarction based on arterial territory involvement. J Clin Neurol 2012; 8: 116–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Glass EN, Cornetta AM, deLahunta A, et al. Clinical and clinicopathologic features in 11 cats with Cuterebra larvae myiasis of the central nervous system. J Vet Intern Med 1998; 12: 365–368. [DOI] [PubMed] [Google Scholar]
- 32. Williams KJ, Summers BA, de Lahunta A. Cerebrospinal cuterebriasis in cats and its association with feline ischemic encephalopathy. Vet Pathol 1998; 35: 330–343. [DOI] [PubMed] [Google Scholar]



