Abstract
Objectives
The aim of this study was to evaluate if a combination of discrete clinical characteristics can be used to identify the most likely differential diagnoses in cats with spinal disease.
Methods
Two hundred and twenty-one cats referred for further evaluation of spinal disease were included and categorised as follows: non-lymphoid neoplasia (n = 44); intervertebral disc disease (n = 42); fracture/luxation (n = 34); ischaemic myelopathy (n = 22); feline infectious peritonitis virus myelitis (n = 18); lymphoma (n = 16); thoracic vertebral canal stenosis (n = 11); acute non-compressive nucleus pulposus extrusion (n = 11); traumatic spinal cord contusion (n = 8); spinal arachnoid diverticula (n = 7); lumbosacral stenosis (n = 5); and spinal empyema (n = 3). Information retrieved from the medical records included signalment, clinical history and clinical presentation. Univariate analyses of variables (clinical history, breed, age, sex, general physical examination findings, onset, progression, spinal hyperaesthesia, asymmetry, ambulatory status and neuroanatomical location) were performed, and variables were retained in a multivariate logistic regression model if P <0.05.
Results
Multivariate logistic regression revealed that intervertebral disc disease most often occurred in middle-aged, purebred cats with a normal general physical examination and an acute onset of painful and progressive clinical signs. Ischaemic myelopathy occurred most often in older cats with a stable or improving, non-painful, lateralising, C6–T2 myelopathy. Spinal fracture/luxation occurred most often in younger cats and resulted most often in a peracute onset, painful, non-ambulatory neurological status. Concurrent systemic abnormalities or abnormal findings detected on general physical examination were significantly associated with feline infectious peritonitis virus myelitis, spinal lymphoma or spinal empyema.
Conclusions and relevance
This study suggests that using easily identifiable characteristics from the history and clinical examination can assist in obtaining a preliminary differential diagnosis when evaluating cats with spinal disease. This information could aid veterinary practitioners in clinical decision-making.
Keywords
Introduction
Assessment of cats with suspected spinal disease can be daunting for veterinary practitioners. Neurophobia is the fear of neuroscience and clinical neurology, which was first recognised in medical students and young physicians. 1 It is associated with the belief that neurology is a complex subject that is academically challenging and difficult to apply in clinical practice. It results in decreased confidence and the inability to apply basic knowledge into clinical practice, essentially leading to paralysis of analysis or ‘paralysis of thinking’.1–3 Following a surge in veterinary neurology research in the last 15 years, neurological diseases are more frequently recognised. Despite the accompanying rise in understanding of neurological disorders, the ‘neurophobia’ phenomenon remains prevalent, particularly among young veterinarians.4,5
A variety of spinal disorders has been recognised in cats, which are associated with different diagnostic approaches, treatment options and varying prognoses.6–8 Infectious disorders, specifically feline infectious peritonitis (FIP) virus myelitis, has historically been considered the most common feline spinal disorder, followed by neoplastic disease, primarily lymphoma.6,8 Other commonly diagnosed feline spinal disorders are spinal fracture and luxation, intervertebral disc disease and ischaemic myelopathy.6–8 With such an extensive list of differential diagnoses, it is not surprising that cats with spinal disease are commonly referred to neurology specialists. Advanced diagnostic tests commonly performed in the referral setting, such as MRI and cerebrospinal fluid (CSF) collection, can however, be invasive and financially prohibitive. 9 Furthermore, not all cat owners will be able or prepared to accept referral to neurology specialists. It is therefore necessary for veterinarians to apply their knowledge and clinical reasoning skills to obtain a likely diagnosis and subsequently consider the necessity, specific advantages, expectations and limitations of a potential referral to a specialist centre.
By considering the signalment, obtaining a thorough clinical history, performing a general physical examination, a complete neurological examination and obtaining a neuroanatomical localisation it is possible to identify key factors that can be vital in clinical decision-making.9–11 It has been identified that most canine spinal diseases are statistically associated with distinct characteristic combinations of clinical variables. 9 However, it is currently unknown if such a statistical model could also be used to guide a clinical reasoning approach in feline spinal disease. The aim of this study was therefore to evaluate if discrete clinical characteristics, such as clinical history, general physical examination findings, signalment, onset, progression, symmetry of clinical signs, spinal hyperaesthesia, ambulatory status and neuroanatomical localisation, can be used to statistically predict the most likely differential diagnoses in cats with spinal disease. We hypothesised that statistical models could be used to identify associations between the most common feline spinal disorders and specific combinations of clinical variables. This information could aid in determining the most likely differential diagnoses when assessing cats with spinal disease and hence improve clinical decision-making for veterinary practitioners.
Materials and methods
This retrospective study was approved by the ethics and welfare committee of the Royal Veterinary College (RVC; SR2018-1663). The digital medical database of the small animal referral hospital, RVC, was searched for all records of cats referred for further evaluation of suspected spinal disease between 2 August 2007 and 3 January 2018. Cats were included if they underwent a complete neurological examination and appropriate further diagnostics to obtain a definitive or presumptive diagnosis of an underlying spinal condition. Further diagnostics could include one or a combination of the following: spinal radiographs, CT, MRI, CSF analysis, infectious disease testing, cytology or histopathology. Cats with sacrocaudal luxation were not included in this study. Cats were excluded if the medical records or imaging studies were incomplete or not available for review, or if a final clinical or presumptive diagnosis was not reached. Although cats were only included if they presented for further evaluation of spinal disease, they were not excluded if the neurological examination revealed abnormalities suggestive of intracranial involvement.All medical records and imaging studies were reviewed by a board-certified neurologist (SDD) and cats were allocated to one of the following 12 disease categories: presumptive non-lymphoid spinal neoplasia; degenerative intervertebral disc disease; spinal fracture/luxation; ischaemic myelopathy; FIP virus myelitis; spinal lymphoma; thoracic vertebral canal stenosis; acute non-compressive nucleus pulposus extrusion; traumatic spinal cord contusion; spinal arachnoid diverticula; lumbosacral stenosis; and spinal empyema. Cases were grouped into a disease category when a diagnosis was made in more than two cats. Cats that suffered from spinal conditions that occurred only once or twice in the study period were therefore not included in this study.For the purpose of this study, a diagnosis of FIP was made when a diagnosis was confirmed by histopathology or detection of feline coronavirus in CSF by real-time reverse transcriptase PCR.12,13 A diagnosis of lymphoma was made by histopathological evaluation or when MRI was suggestive for a neoplastic spinal condition and cytological evaluation of CSF or extraneural tissue was suggestive for lymphoma. 14 A diagnosis of presumptive non-lymphoid spinal neoplasia was made when neoplastic disease other than lymphoma was histopathologically confirmed or if cytological evaluation of CSF or extraneural tissues did not reveal any indications of lymphoma. A diagnosis of thoracic vertebral canal stenosis was defined as a focal osseous vertebral canal stenosis. Diagnostic criteria for the other spinal disease categories were based on previously published literature.15–23
For all included cases, the following information was retrieved from the medical records: clinical history with an emphasis on the occurrence of other clinical signs such as lethargy, anorexia and weight loss; signalment; onset, duration, type and severity of clinical signs; and general physical and neurological examination findings, including lateralisation of clinical signs and presence of spinal hyperaesthesia. Age was classified as younger (<3 years), middle-aged (3–9 years) and older (>9 years). Onset of clinical signs was categorised into peracute (<2 days), acute (2–5 days), subacute (5–14 days) and chronic (>14 days). Progression of clinical signs was categorised into deteriorating, static or improving clinical signs before presentation at the RVC. This assessment was based on the notes from the referring veterinary surgeon and the owner’s perception. Severity of clinical signs was categorised into ambulatory or non-ambulatory neurological status on presentation. Spinal hyperaesthesia was considered to be present when a painful response could be elicited on spinal palpation by the attending clinician or when obvious spinal pain was reported by the referring veterinary surgeon or owner of the cat. Neurological signs were considered to be lateralised when there was an unequivocal difference in the severity of neurological deficits between the left and right side of the cat. The neuroanatomical localisation was categorised into disorders affecting the C1–C5, C6–T2, T3–L3 or L4–S3 spinal cord segments or multifocal with intracranial involvement.
CT was performed with a 16-slice helical CT scanner (PQ 500, Universal Systems, Solon; GE Healthcare), under sedation or general anaesthesia. After completion of the transverse CT study, sagittal, dorsal and three-dimensional reconstructions were made. MRI was performed under general anaesthesia with a high-field unit (1.5 T, Intera; Philips Medical Systems) and imaging studies included a minimum of T1- and T2-weighted sagittal and transverse images.
Statistical analysis was performed by one of the authors (TAJC) and data were analysed using statistical software (SPSS V.21.0.1; IBM).
Univariate analyses of potential explanatory variables for each condition were performed. Variables were considered for inclusion in multivariate logistic regression if P <0.30 and were retained in the final model if P <0.05, based on the likelihood ratio test. Multivariate logistic regression was carried out using a forced entry method (where all variables are entered into the equation in a single step) to examine associations between included variables with a significance level of P <0.05. 24 Results are presented with odds ratios (ORs) and 95 percent confidence intervals (CIs) for each condition vs the overall spinal disease population. 24 Following multivariate logistic regression for each disease variables retained in the final model (P <0.05) included purebred status, age (signalment), concurrent abnormalities in the clinical history or general physical examination, median time to presentation, progression of clinical signs, ambulatory status, spinal hyperaesthesia, asymmetry in neurological deficits and neuroanatomical localisation. Non-normally distributed data are presented as median (range). Normally distributed data are presented as mean ± SD.
Results
Two hundred and twenty-six cats were diagnosed with a spinal condition in the study period. Five cats were excluded because their diagnosis occurred only once or twice. These five cats were diagnosed with traumatic intramedullary haemorrhage (n = 2), suspected poliomyelitis, toxoplasmosis (n = 1) and vertebral malformation caused by mucopolysaccharidosis (n = 1).
Two hundred and twenty-one cats were therefore included in this study. This group consisted of 143 males (131 neutered) and 78 females (67 neutered) aged between 2 months and 18 years. The most commonly diagnosed condition was presumptive non-lymphoid neoplasia (n = 44; 19.9%), followed by degenerative intervertebral disc disease (n = 42; 19.0%), spinal fracture and luxation (n = 34; 15.4%), ischaemic myelopathy (n = 22; 10.0%), FIP virus myelitis (n = 18; 8.1%), lymphoma (n = 16; 7.2%), thoracic vertebral canal stenosis (n = 11; 5.0%), acute non-compressive nucleus pulposus extrusion (n = 11; 5.0%), traumatic spinal cord contusion (n = 8; 3.6%), spinal arachnoid diverticulum (n = 7; 3.2%), lumbosacral stenosis (n = 5; 2.3%) and spinal empyema (n = 3; 1.4%). A summary of the clinical presentation of cats affected by these disorders is presented in Table 1.
Prevalence and clinical characteristics of 221 cats with spinal disease
P = peracute; A = acute; S = subacute; C = chronic; D = deteriorating; S = static; Imp = improving
The 44 cats with presumptive non-lymphoid neoplasia included 15 cats with contrast-enhancing intramedullary mass lesions. Serum toxoplasma titres were negative and CSF analysis was within normal limits in all these 15 cats. Thirteen cats had vertebral masses, of which five were histopathologically confirmed to be osteosarcoma and one was confirmed to be a plasmacytoma; six cats had histopathologically confirmed meningioma; two cats had histopathologically confirmed glial cell tumours; two had unspecified extradural mass lesions; and each of the following diagnoses were made in one cat: vascular hamartoma, fibrosarcoma, solitary giant cell tumour of soft tissue, histiocytic sarcoma, peripheral nerve sheath tumour and metastatic neoplasia.
Age
Older age was associated with a diagnosis of presumptive non-lymphoid neoplasia, ischaemic myelopathy and lumbosacral stenosis (Table 2). Cats with degenerative intervertebral disc disease were more likely middle-aged and cats with spinal fracture and luxation, FIP virus myelitis and traumatic spinal cord contusion were more likely to be younger (Table 2).
Multivariate logistic regression analysis of signalment, clinical presentation and clinical examination characteristics of feline spinal disorders with more than two cases
Where statistically significant (P ⩽0.05), data presented include odds ratio with 95% confidence intervals indicated in parentheses
Breed
In total, 33% of cats in this study were purebred (n = 55) and 67% were non-purebred (n = 166). The group of non-purebred cats consisted of domestic shorthair (n = 143 cats), domestic longhair (n = 19) and domestic mediumhair cats (n = 4). The most common purebred cat was the British Shorthair (n = 11), followed by the Bengal (n = 9), Persian (n = 8), Maine Coon (n = 7), Sphynx and Siamese (n = 3 for both), Russian Blue, Chinchilla, Tonkinese and Ragdoll (n = 2 for each) and six breeds were represented by only one cat. Purebred status was significantly associated with a diagnosis of presumptive non-lymphoid neoplasia, degenerative intervertebral disc disease and thoracic vertebral canal stenosis. Cats with degenerative intervertebral disc disease and thoracic vertebral canal stenosis were more likely to be purebred cats, while cats with presumptive non-lymphoid neoplasia were more likely to be non-purebred cats (Table 2).
Concurrent clinical signs and general physical examination findings
Compared with other diagnoses, cats with FIP virus myelitis, lymphoma and spinal empyema more often had concurrent clinical signs, such as lethargy, anorexia and weight loss, or abnormalities on their general physical examination, such as pyrexia and lymphadenomegaly. Cats with degenerative intervertebral disc disease had concurrent clinical signs or abnormalities significantly less often on their general physical examination (Table 2).
Onset and progression of clinical signs
Onset of disease was significantly associated with diagnoses of degenerative intervertebral disc disease and vertebral fracture and luxation. Cats with degenerative intervertebral disc disease more likely had an acute onset of clinical signs, while cats with vertebral fracture and luxation were more likely to have a peracute onset of clinical signs (Table 2). Progression of clinical signs was significantly associated with diagnoses of presumptive non-lymphoid neoplasia, degenerative intervertebral disc disease and ischaemic myelopathy. Cats with presumptive non-lymphoid neoplasia and degenerative intervertebral disc disease had more likely deteriorating clinical signs, while cats with ischaemic myelopathy were more likely to demonstrate static or improving clinical signs (Table 2).
Neurological examination findings
Neuroanatomical localisation
The neuroanatomical localisation was significantly associated with diagnoses of ischaemic myelopathy and FIP virus myelitis. Cats with ischaemic myelopathy were more likely to have a lesion localised to the C6–T2 spinal cord segments, while cats with FIP virus myelitis were more likely to have a multifocal neuroanatomical localisation with intracranial involvement (Table 2).
Ambulatory status
Ambulatory status was significantly associated with diagnoses of spinal fracture and luxation and acute non-compressive nucleus pulposus extrusion. Cats with spinal fracture and luxation or acute non-compressive nucleus pulposus extrusion were more likely to not be ambulatory at the time of presentation (Table 2).
Presence of spinal hyperaesthesia
The presence of spinal hyperaesthesia was significantly associated with diagnoses of degenerative intervertebral disc disease, spinal fracture and luxation, ischaemic myelopathy and thoracic vertebral canal stenosis. Cats with degenerative intervertebral disc disease, spinal fracture and luxation, and thoracic vertebral canal stenosis were more likely to demonstrate spinal hyperaesthesia, while cats with ischaemic myelopathy were less likely to demonstrate spinal hyperaesthesia (Table 2).
Lateralisation of clinical signs
The presence of obviously lateralised clinical signs was significantly associated with diagnoses of ischaemic myelopathy and thoracic vertebral canal stenosis. Cats with ischaemic myelopathy and thoracic vertebral canal stenosis were more likely to demonstrate lateralisation of their clinical signs (Table 2).
Discussion
This study evaluated whether discrete clinical characteristics can be used to help in identifying the most likely differential diagnoses in cats with spinal disease. Our results suggest that the most common feline spinal disorders are statistically associated with discrete variables obtained from the clinical history, signalment, and general physical and neurological examinations. Owing to the extensive list of possible diagnoses and the associated variation in prognoses of cats with spinal disease, achieving a ‘most likely’ differential diagnosis before carrying out further diagnostics is invaluable, particularly in the first-opinion setting, where finances can be a major concern.In agreement with our findings, previous studies evaluating canine spinal disease and canine and feline epilepsy highlighted how problem-based clinical reasoning enabled a diagnostic process that was focused at the level of the signalment, history, clinical signs and neurological examination.9–11 Clinical reasoning can be considered a thinking process in which we collect and process multiple fragments of clinical information, come to an understanding of a patient’s clinical problem, and use this integrated information to plan further diagnostic and therapeutic interventions. Following this approach can help to break down complex and potentially overwhelming clinical presentations into logical and manageable cases. 25 We therefore hope that the results of this study will improve clinical decision-making for veterinary surgeons managing cats with spinal disease.
The most common feline spinal disorders in this study were presumptive non-lymphoid neoplasia, followed by intervertebral disc disease, fracture and luxation, and ischaemic myelopathy. FIP virus myelitis was only the fifth most common spinal disorder. This finding is different from previous data suggesting that FIP virus myelitis should be considered the most common spinal disorder in cats.6,8 This difference can potentially be explained by geographical differences in the prevalence of spinal disorders and infectious diseases in particular. Another contributing factor could be the different inclusion criteria used in studies. A previous study evaluating the prevalence of spinal disorders in cats included cases for which a histopathological diagnosis was available. 6 Although this inclusion criterion has the clear advantage that only cases with a definitive diagnosis were included, a histopathological diagnosis is typically only obtained after completion of a necropsy. This inclusion criterion could therefore potentially favour the selection of cases with a poor prognosis, such as FIP virus myelitis and spinal neoplasia.It should further be emphasised that our study only included cats that presented for further evaluation of spinal disease. Although we did not exclude cases for which the neurological examination revealed abnormalities suggestive for intracranial involvement, we did not include cats for which spinal disease was part of a more complex and multifocal neurological presentation. Although it is possible that our study therefore represents a more accurate reflection of the prevalence of feline spinal disorders in a referral clinical setting, a major limitation is the lack of a definitive diagnosis in several cases. This is especially true for the group of non-lymphoid spinal tumours, which was considered the most common diagnosis in our study. This diagnosis was more common in older, non-purebred cats with deteriorating clinical signs.
For the purpose of this study, we grouped cats with spinal lymphoma into a separate disease category. The reasons for this were that spinal lymphoma has historically been considered one of the most common feline spinal disorders and that spinal lymphoma has been associated with different clinical characteristics compared with other feline spinal tumours. Lymphoma has been suggested to be the most common spinal tumour in cats, representing up to 39% of spinal tumours in this species. 26 Compared with cats with other spinal tumours, cats with lymphoma have been suggested to be younger, have a more rapid progression of clinical signs, have more often lateralised or asymmetrical neurological deficits, and have more often clinical signs localised to the thoracic or lumbosacral spinal segments.26–28 However, our results suggest that it is difficult to differentiate lymphoma from other feline spinal disorders without further diagnostics. The only clinical variable significantly associated with a diagnosis of spinal lymphoma was the presence of concurrent clinical signs and abnormalities on general physical examination. These findings are in agreement with previous suggestions that spinal lymphoma may be difficult to differentiate from other spinal disorders and that non-specific signs such as anorexia, lethargy and weight loss commonly precede neurological signs. 29 It is well known that some common feline neurological conditions are expressions of systemic disease, which is illustrated by the fact that lymphoma, FIP virus myelitis and spinal empyema were significantly associated with concurrent clinical signs and abnormalities on general physical examination. The presence of such abnormalities was associated with more than 30 times the odds for the diagnoses of spinal lymphoma and FIP virus myelitis. A diagnosis of FIP virus myelitis was further associated with a young age and a multifocal neuroanatomical localisation, which is in agreement with previous studies.12,13
Although the prevalence of degenerative intervertebral disc disease in the overall feline population should be considered low,20,30 this was the second most common spinal disorder in our study. This condition was significantly associated with middle-aged, purebred cats with no abnormalities detected on general physical examination that developed an acute onset of progressive and painful clinical signs (Table 2). These findings are in agreement with previous studies that have reported spinal hyperaesthesia and progressive clinical signs in the majority of cases,20,30,31 and have suggested that purebred cats, in particular Persians and British Short-hairs, are predisposed for intervertebral disc disease. 20 Previous studies have also suggested that most cats are young to middle-aged, 31 with a mean age at time of diagnosis ranging from 9.5 to 9.8 years.20,30
In agreement with previous findings, spinal fracture and luxation was a common cause of spinal disease in this study.6,15 This is not surprising given the partial outdoor lifestyle of most cats. This condition was associated with young cats that presented with a peracute onset of a non-ambulatory neurological status and spinal hyperaesthesia. Spinal fracture and luxation can be considered a severe spinal emergency in cats. Surgical treatment is technically challenging, expensive and can be associated with an uncertain prognosis.15,17,32,33 It is important to realise that cats that are involved in a traumatic incident can also suffer from other spinal conditions. Acute non-compressive nucleus pulposus extrusion and spinal cord contusion, two conditions often associated with external trauma, were also considered common spinal conditions in this study.19,21 Treatment of both conditions does not involve surgery, and this illustrates that multiple differential diagnoses should be considered when a cat is presented after suspected spinal trauma.
Ischaemic myelopathy was the fourth most common feline spinal disorder and was, in agreement with previous studies, associated with a characteristic clinical presentation. Cats with ischaemic myelopathy were typically older and presented with stable or improving, non-painful, lateralised clinical signs.16,34 The presence of improving clinical signs was considered the strongest clinical indicator for a diagnosis of ischaemic myelopathy (Table 2). This condition was also associated with a C6–T2 neuroanatomical localisation, which is in agreement with previous findings. 16
The main limitations of this study were its retrospective study design and the inclusion of cases without a histopathologically confirmed diagnosis. Although for most disease categories a diagnosis was based on previously published criteria and a board-certified neurologist reviewed all diagnostic studies, it is possible that some cases might have been incorrectly classified. It is possible that this methodology enabled inclusion of disorders with a more favourable prognosis and provided, therefore, a more accurate reflection of the overall caseload seen in a tertiary referral population. It should, however, also be emphasised that all included cats were, indeed, referred to a specialist referral hospital and all underwent advanced diagnostics. It is therefore possible that the prevalence of spinal disorders reported in this study cannot be reliably extrapolated to a first-opinion setting. It is possible that easy-to-diagnose spinal conditions, such as spinal fracture/luxation, and conditions with mild clinical signs are less likely referred for further evaluation by specialists. It should further be emphasised that cats with sacrocaudal luxation or ‘tail-pull injury’ were not included in this study. Although this is a commonly encountered condition, sacrocaudal luxation is associated with specific clinical characteristics, 35 which can be considered distinct from those of cats suffering from ‘other’ spinal disease.
Conclusions
Variables from the clinical history, signalment, and general physical and neurological examinations can be systematically evaluated to construct a focused and prioritised list of differential diagnoses, allowing the implementation of an appropriate diagnostic and treatment approach. Not only does this help with guiding clients and their expectations, but it can also help clinicians by increasing their confidence and decreasing stress when evaluating cats with suspected spinal disease.
Footnotes
Author note
The results of this study were presented in abstract form (poster presentation) at the 31st symposium of the European Society of Veterinary Neurology – European College of Veterinary Neurology (ESVN-ECVN), 20–22 September 2018, Copenhagen, Denmark.
Conflict of interest
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.
Ethical approval
This work involved the use of non-experimental animals only (owned or unowned), and followed established internationlly recognised high standards (‘best practice’) of individual veterinary clinical patient care. Ethical approval from a committee was not necessarily required.
Informed consent
Informed consent (either verbal or written) was obtained from the owner or legal custodian of all animal(s) described in this work for the procedure(s) undertaken. No animals or humans are identifiable within this publication, and therefore additional informed consent for publication was not required.
