Cerebrospinal Fluid Analysis in Small Animal Neurology
By Dr. Zubair Khalid, DVM, MS, PhD ·
Key Takeaways
- Cerebrospinal fluid (CSF) analysis is crucial for diagnosing central nervous system (CNS) diseases in small animals, reflecting metabolic, inflammatory, infectious, neoplastic, and degenerative processes.
- Pleocytosis, an elevated total nucleated cell count (TNCC) above 5 cells/µL, is a key indicator of CNS inflammation, with neutrophilic pleocytosis often pointing to bacterial meningitis or steroid-responsive meningitis-arteritis (SRMA), while mixed or mononuclear pleocytosis suggests meningoencephalitis of unknown origin (MUO) or protozoal infections.
- Albuminocytological dissociation, characterized by elevated CSF protein without pleocytosis, is strongly associated with compressive or degenerative spinal cord diseases such as intervertebral disc herniation or spinal neoplasia.
- Infectious etiologies are investigated using serological tests, antigen detection (e.g., ELISA), and molecular methods like polymerase chain reaction (PCR) or reverse transcription PCR (RT-PCR) for agents such as Toxoplasma gondii, Neospora caninum, and feline coronavirus.
- Differentiating MUO from infectious meningoencephalitis requires integrating CSF findings with neuroimaging (MRI) and comprehensive infectious disease testing, as CSF analysis alone is often insufficient.
- Iatrogenic blood contamination is a common artifact that can confound interpretation of cell counts and protein levels; while correction factors exist, they are imprecise, and immediate sample processing is vital to prevent cellular degeneration.
Introduction
Cerebrospinal fluid (CSF) analysis is a cornerstone diagnostic procedure in small animal neurology, providing critical information about the pathophysiological state of the central nervous system (CNS). CSF bathes the brain and spinal cord, and its composition directly reflects metabolic, inflammatory, infectious, neoplastic, and degenerative processes within the neuraxis [<a href="#ref-1">1</a>, <a href="#ref-2">2</a>]. Collection and analysis of CSF are indicated whenever CNS disease is suspected based on neurological examination, particularly when inflammatory or infectious etiologies are considered [<a href="#ref-3">3</a>, <a href="#ref-4">4</a>]. The diagnostic yield of CSF analysis depends on appropriate collection technique, timely processing, and careful interpretation of cellular and biochemical parameters within the context of neuroimaging findings and clinical presentation [<a href="#ref-2">2</a>, <a href="#ref-5">5</a>].
Collection Techniques and Procedural Considerations
CSF is most commonly collected from the cerebellomedullary cistern (cisternal puncture) or the lumbar subarachnoid space (lumbar puncture) in dogs and cats. General anesthesia is required to minimize patient movement, reduce the risk of iatrogenic trauma, and facilitate proper positioning [<a href="#ref-6">6</a>]. Cisternal puncture typically yields higher cell counts and protein concentrations compared to lumbar collection because of the rostrocaudal gradient of CSF flow [<a href="#ref-2">2</a>]. The volume collected should be limited to 0.5 to 1 mL per 5 kg of body weight to avoid complications from CSF depletion.
Iatrogenic blood contamination is the most common artifact in CSF analysis and can confound interpretation. Blood contamination can result from trauma to the dorsal vertebral venous plexus during needle placement [<a href="#ref-6">6</a>]. The presence of subarachnoid hemorrhage or axonal degeneration has been described after cervical centesis in horses, and similar risks exist in small animals, particularly when needle placement is difficult [<a href="#ref-6">6</a>]. Samples should be collected into sterile tubes containing EDTA for cytological evaluation and into plain tubes for biochemical analysis. Immediate processing within 30 to 60 minutes is recommended because cellular degeneration, particularly of neutrophils, occurs rapidly in CSF samples [<a href="#ref-2">2</a>].
Cellular Analysis
Total Nucleated Cell Count
Total nucleated cell count (TNCC) is a fundamental parameter in CSF analysis. Reference intervals for dogs and cats typically range from 0 to 5 cells per microliter, with variations depending on collection site and laboratory methodology [<a href="#ref-2">2</a>]. Automated impedance analyzers and hemocytometer chambers are used for enumeration. Cell counts above 5 cells per microliter are considered pleocytosis. Pleocytosis can be classified as mild (5 to 50 cells/microliter), moderate (50 to 500 cells/microliter), or severe (greater than 500 cells/microliter). The degree of pleocytosis correlates with disease severity but not with specific etiology in all cases [<a href="#ref-2">2</a>, <a href="#ref-7">7</a>].
Differential Cell Count and Cytological Patterns
Cytological evaluation of CSF sediment after concentration by cytocentrifugation allows identification of cell types and recognition of pathological patterns. The normal CSF contains predominantly small mononuclear cells, with occasional lymphocytes and rare monocytes. Neutrophils and eosinophils are absent in normal CSF [<a href="#ref-2">2</a>]. The proportions of cell types shift in disease states, and recognition of predominant patterns guides differential diagnosis.
Neutrophilic pleocytosis is characterized by an increased proportion of nondegenerate neutrophils. This pattern is most commonly associated with bacterial meningitis, fungal infections, and steroid-responsive meningitis-arteritis (SRMA) [<a href="#ref-8">8</a>, <a href="#ref-9">9</a>]. In SRMA, neutrophilic pleocytosis may be severe and accompanied by elevated protein concentration [<a href="#ref-8">8</a>]. In bacterial infections, neutrophils may show degenerative changes such as karyolysis and cytoplasmic vacuolation. However, neutrophilic pleocytosis is not specific for infection; it can also occur in meningoencephalitis of unknown origin (MUO) and following intracranial surgery or trauma [<a href="#ref-8">8</a>].
Mixed cell pleocytosis involves increases in both neutrophils and mononuclear cells. This pattern is typical of granulomatous meningoencephalomyelitis (GME), necrotizing meningoencephalitis (NME), and other noninfectious inflammatory diseases collectively classified as MUO [<a href="#ref-3">3</a>, <a href="#ref-10">10</a>]. In dogs with MUO, CSF analysis frequently reveals mononuclear or mixed pleocytosis with moderately elevated protein concentration [<a href="#ref-3">3</a>, <a href="#ref-10">10</a>]. Distinguishing MUO from infectious meningoencephalitis requires integration of CSF findings with serological, molecular, and neuroimaging data [<a href="#ref-3">3</a>].
Eosinophilic pleocytosis is defined by an eosinophil proportion exceeding 10% of the total nucleated cells. This pattern is rare but has been associated with protozoal infections (Neospora caninum, Toxoplasma gondii), fungal infections, parasitic migration, and primary CNS histiocytic sarcoma [<a href="#ref-9">9</a>, <a href="#ref-11">11</a>, <a href="#ref-12">12</a>]. In one reported case, persistent marked CSF eosinophilia was documented in a dog with primary CNS histiocytic sarcoma, demonstrating that neoplasia should be considered in the differential diagnosis of eosinophilic pleocytosis [<a href="#ref-11">11</a>]. Nematode eggs have been observed in CSF cytology from dogs with intramedullary Spirocerca lupi spinal cord migration, establishing a specific parasitic etiology for eosinophilic inflammation [<a href="#ref-12">12</a>].
Mononuclear pleocytosis with predominantly lymphocytes and macrophages is observed in viral encephalitides, including canine distemper virus infection, and in some cases of MUO [<a href="#ref-2">2</a>, <a href="#ref-3">3</a>]. In feline infectious peritonitis (FIP), CSF analysis often reveals elevated protein concentration with a mixed or mononuclear pleocytosis, although the diagnostic hallmark is the detection of coronavirus antigen or antibody in CSF [<a href="#ref-13">13</a>].
Red Blood Cell Count and Blood Contamination
The presence of erythrocytes in CSF may indicate iatrogenic blood contamination or pathological hemorrhage. A red blood cell count exceeding 500 cells per microliter suggests either traumatic tap or subarachnoid hemorrhage [<a href="#ref-14">14</a>]. In dogs with suspected cerebrovascular disease, xanthochromia (yellowish discoloration of the supernatant after centrifugation) supports a diagnosis of prior hemorrhage [<a href="#ref-14">14</a>]. Iatrogenic blood contamination complicates interpretation of protein concentration and cell counts. A commonly used correction factor subtracts one nucleated cell and one milligram of protein per 1000 red blood cells, although this correction is imprecise and should be applied cautiously [<a href="#ref-6">6</a>].
Biochemical Analysis
Protein Concentration
Total protein concentration in CSF is typically measured using turbidimetric or colorimetric methods. Reference intervals for cisternal CSF range from 15 to 40 mg/dL in dogs and less than 30 mg/dL in cats. Lumbar CSF generally has a higher protein concentration than cisternal CSF [<a href="#ref-2">2</a>]. Elevations in CSF protein occur due to increased blood-brain barrier (BBB) permeability, intrathecal immunoglobulin production, or impaired CSF absorption.
Albuminocytological Dissociation
Albuminocytological dissociation refers to elevated CSF protein concentration in the absence of pleocytosis. A large retrospective study of CSF analysis in dogs found that albuminocytological dissociation occurred in approximately 12% of samples [<a href="#ref-2">2</a>]. This pattern is most commonly associated with compressive or degenerative spinal cord diseases, including intervertebral disc herniation, spinal neoplasia, and degenerative myelopathy [<a href="#ref-2">2</a>, <a href="#ref-7">7</a>]. In dogs without deep pain perception due to thoracolumbar disc herniation, elevated CSF protein has been evaluated as a prognostic indicator [<a href="#ref-7">7</a>].
CSF Lactate
CSF lactate concentration reflects anaerobic metabolism within the CNS. Lactate is produced by glycolysis in neurons and glial cells under hypoxic conditions and is normally cleared by the CSF circulation. In dogs with CNS disease, CSF lactate concentration has been associated with seizure activity, cerebral ischemia, and inflammation [<a href="#ref-15">15</a>]. However, lactate measurement has limited specificity for differentiating infectious from noninfectious etiologies [<a href="#ref-15">15</a>].
Oligoclonal Bands and Immunoglobulin Analysis
Oligoclonal bands (OCBs) are discrete bands of immunoglobulin G (IgG) detected by isoelectric focusing or electrophoresis of CSF and matched serum samples. The presence of OCBs in CSF without corresponding bands in serum indicates intrathecal IgG synthesis. In human neurology, OCBs are a hallmark of multiple sclerosis. In dogs, OCBs have been detected in a subset of dogs with idiopathic epilepsy, suggesting possible immune-mediated mechanisms in certain epilepsy phenotypes [<a href="#ref-16">16</a>]. The utility of OCB detection in differentiating MUO from infectious meningoencephalitis remains under investigation [<a href="#ref-16">16</a>].
Infectious Disease Testing
Serological and Molecular Pathogen Detection
Detection of infectious agents in CSF can be achieved through serological testing (antibody detection), antigen testing (ELISA), or molecular methods such as polymerase chain reaction (PCR) [<a href="#ref-3">3</a>, <a href="#ref-13">13</a>]. PCR offers high sensitivity and specificity for detecting pathogen nucleic acid in CSF. In dogs with suspected infectious meningoencephalomyelitis, a comprehensive testing panel including serology and PCR for Toxoplasma gondii, Neospora caninum, Cryptococcus neoformans, and canine distemper virus is recommended [<a href="#ref-3">3</a>].
In cats with neurologic FIP, detection of feline coronavirus RNA by reverse transcription PCR (RT-PCR) in CSF is highly specific, while anti-coronavirus antibody detection in CSF supports the diagnosis but has lower specificity [<a href="#ref-13">13</a>]. Fungal infections of the CNS, including cryptococcosis, blastomycosis, and histoplasmosis, can be diagnosed by CSF culture, antigen detection, or cytological identification of organisms [<a href="#ref-9">9</a>]. In one case of neurocandidiasis in a dog, clinicopathological and diagnostic imaging findings were described, highlighting the diagnostic challenges of fungal CNS infections [<a href="#ref-17">17</a>].
Metabolite Biomarkers
Mass spectrometry-based profiling of CSF metabolites has emerged as a diagnostic tool for identifying CNS involvement in infectious diseases. In human patients with varicella zoster virus reactivation, specific metabolite biomarkers in CSF distinguished cases with CNS involvement from those without [<a href="#ref-18">18</a>]. Similar approaches are being explored in veterinary medicine for differentiating infectious from noninfectious CNS diseases [<a href="#ref-1">1</a>, <a href="#ref-18">18</a>].
Diagnostic Patterns and Clinical Correlates
Meningoencephalitis of Unknown Origin
Meningoencephalitis of unknown origin encompasses a group of noninfectious inflammatory CNS diseases, including granulomatous meningoencephalomyelitis (GME), necrotizing meningoencephalitis (NME), and necrotizing leukoencephalitis (NLE). CSF analysis in MUO typically reveals mononuclear or mixed pleocytosis with elevated protein concentration [<a href="#ref-3">3</a>, <a href="#ref-10">10</a>]. In a large study comparing clinical signs, CSF analysis, neuroimaging, and infectious disease testing in 168 dogs, approximately half of the cases were classified as MUO after exclusion of infectious etiologies [<a href="#ref-3">3</a>]. CSF analysis alone cannot reliably distinguish MUO from infectious meningoencephalitis; integration with MRI findings and infectious disease testing is required [<a href="#ref-3">3</a>].
Neospora caninum Meningoencephalitis
Cerebrospinal fluid analysis in dogs with Neospora caninum meningoencephalitis often reveals mixed or mononuclear pleocytosis with elevated protein concentration [<a href="#ref-19">19</a>, <a href="#ref-20">20</a>]. A comparison of serum creatine kinase and aspartate aminotransferase activity in dogs with Neospora meningoencephalitis and noninfectious meningoencephalitis found that these muscle enzyme markers differ between the two groups, providing ancillary diagnostic information [<a href="#ref-20">20</a>]. Masticatory muscle changes detected on MRI also help differentiate Neospora caninum infection from MUO [<a href="#ref-19">19</a>].
Idiopathic Epilepsy
In dogs with idiopathic epilepsy, CSF analysis is typically normal or shows mild abnormalities. A study of CSF-specific oligoclonal bands in dogs with idiopathic epilepsy demonstrated intrathecal IgG synthesis in a subset of cases, suggesting that immune mechanisms may contribute to seizure generation in some individuals [<a href="#ref-16">16</a>]. The International Veterinary Epilepsy Task Force consensus statement recommends CSF analysis as part of the diagnostic workup for epilepsy, particularly when structural epilepsy is suspected based on neurological examination or MRI findings [<a href="#ref-4">4</a>].
Hydrocephalus and CSF Flow Abnormalities
Communicating internal hydrocephalus in small breed dogs may be associated with altered CSF flow dynamics. Phase-contrast MRI has identified low CSF velocity at the foramen magnum in small breed dogs with enlarged ventricular systems [<a href="#ref-21">21</a>]. Intraoperative measurement of intraventricular pressure in dogs with communicating internal hydrocephalus has provided insights into the pathophysiology of this condition and the role of CSF analysis in surgical planning [<a href="#ref-22">22</a>]. CSF flow abnormalities detected by time-spatial labeling inversion pulse MRI have also been observed in small breed dogs with idiopathic epilepsy, suggesting possible overlap between hydrocephalus and epilepsy phenotypes [<a href="#ref-23">23</a>].
Otitis Media and Interna in Cats
In cats with otitis media and interna, the presence of meningeal enhancement on postcontrast MRI may be associated with CSF abnormalities. A retrospective study found that CSF analysis in cats with otitis media and interna with meningeal enhancement more frequently showed pleocytosis and elevated protein compared to cats without meningeal enhancement [<a href="#ref-24">24</a>]. These findings support the use of CSF analysis to assess intracranial extension of middle ear disease.
Biomarker Development and Proteomics
Mass spectrometry-based proteomic and metabolomic profiling of CSF holds promise for identifying novel biomarkers of CNS disease in small animals. A study profiling CSF metabolites in dogs with various neurological disorders identified potential biomarker panels for distinguishing neurodegenerative, inflammatory, and neoplastic conditions [<a href="#ref-1">1</a>]. Biomarkers such as tau protein, amyloid-beta, and neurofilament light chain have been investigated in canine CSF for their diagnostic and prognostic utility [<a href="#ref-1">1</a>].
The following table summarizes the primary CSF findings associated with common disease categories.
| Disease Category | Typical TNCC (cells/microliter) | Predominant Cell Type | Protein Elevation | Key Ancillary Findings |
|---|---|---|---|---|
| Normal | 0 to 5 | Small mononuclear | None | Clear, colorless |
| MUO | 10 to 500 | Mononuclear or mixed | Moderate | Negative infectious testing |
| Bacterial meningitis | 100 to >1000 | Neutrophils (degenerate) | Marked | Positive culture or PCR |
| SRMA | 100 to >1000 | Neutrophils (nondegenerate) | Marked | Negative infectious testing |
| Neospora caninum | 10 to 200 | Mixed or mononuclear | Moderate | Positive serology or PCR |
| FIP (feline) | 10 to 100 | Mixed or mononuclear | Marked | Positive coronavirus RT-PCR |
| Fungal infection | 10 to 500 | Mixed with eosinophils | Moderate to marked | Positive antigen test or culture |
| CNS neoplasia | 0 to 50 | Mononuclear | Mild to moderate | Neoplastic cells (rare) |
| Albuminocytological dissociation | 0 to 5 | Normal | Mild to marked | Compressive or degenerative disease |
Diagnostic Workflow
The following Mermaid diagram illustrates a recommended diagnostic workflow for CSF analysis in small animal neurology.
flowchart TD
A["Neurological Examination"] --> B{"CSF Collection Indicated?"}
B -->|"Yes"| C["Perform Cisternal or Lumbar Puncture"]
B -->|"No"| D["Alternative Diagnostics / Monitoring"]
C --> E["Gross Assessment: Color, Clarity"]
E --> F["Total Nucleated Cell Count and RBC Count"]
F --> G["Biochemical Analysis: Protein, Lactate"]
G --> H["Concentration by Cytocentrifugation"]
H --> I["Cytological Evaluation and Differential Count"]
I --> J{"Pattern Recognition"}
J -->|"Neutrophilic"| K["Infectious / SRMA / Post-Surgical"]
J -->|"Mixed Mononuclear/Neutrophilic"| L["MUO / Protozoal / Fungal"]
J -->|"Mononuclear"| M["Viral / MUO / Neoplastic"]
J -->|"Eosinophilic"| N["Parasitic / Fungal / Protozoal / Neoplastic"]
J -->|"Normal with Elevated Protein"| O["Albuminocytological Dissociation"]
K & L & M & N & O --> P["Infectious Disease Testing (PCR, Serology)"]
P --> Q["Integrate with MRI and Clinical Data"]
Q --> R["Final Diagnosis and Treatment Plan"]Frequently Asked Questions
What is the normal total nucleated cell count in canine CSF?
The normal TNCC in canine cisternal CSF is 0 to 5 cells per microliter, with lumbar CSF typically having a slightly higher count [<a href="#ref-2">2</a>].
What does albuminocytological dissociation indicate?
Albuminocytological dissociation indicates elevated CSF protein without pleocytosis and is most commonly associated with compressive spinal cord diseases, degenerative myelopathy, or spinal neoplasia [<a href="#ref-2">2</a>, <a href="#ref-7">7</a>].
When is CSF eosinophilia clinically significant?
CSF eosinophilia is significant when eosinophils exceed 10% of the total nucleated cells, suggesting parasitic migration, protozoal infection, fungal disease, or rarely CNS neoplasia such as histiocytic sarcoma [<a href="#ref-9">9</a>, <a href="#ref-11">11</a>, <a href="#ref-12">12</a>].
Can CSF analysis distinguish MUO from infectious meningoencephalitis?
CSF analysis alone cannot definitively distinguish MUO from infectious meningoencephalitis; integration with MRI findings and comprehensive infectious disease testing is required [<a href="#ref-3">3</a>].
Is blood-contaminated CSF still diagnostically useful?
Blood-contaminated CSF remains useful for culture, PCR, and certain biochemical analyses, but cytological interpretation is compromised. Correction formulas for cell count and protein are imprecise and should be applied cautiously [<a href="#ref-2">2</a>, <a href="#ref-6">6</a>].
What is the role of CSF lactate in neurological diagnosis?
CSF lactate concentration reflects CNS anaerobic metabolism and may be elevated in seizures, ischemia, and inflammation, but it has limited specificity for differentiating disease etiologies [<a href="#ref-15">15</a>].
Are oligoclonal bands detected in canine CSF?
Yes, oligoclonal bands have been detected in the CSF of dogs with idiopathic epilepsy, suggesting intrathecal IgG synthesis in a subset of cases [<a href="#ref-16">16</a>].
What CSF abnormalities are expected in feline infectious peritonitis?
CSF analysis in cats with neurologic FIP typically reveals elevated protein concentration with mild to moderate pleocytosis; detection of coronavirus RNA by RT-PCR is highly specific [<a href="#ref-13">13</a>].
How does CSF analysis guide therapy in steroid-responsive meningitis-arteritis?
In SRMA, CSF analysis typically shows marked neutrophilic pleocytosis and elevated protein. Response to corticosteroid therapy is monitored by serial CSF analysis, with normalization of parameters indicating effective treatment [<a href="#ref-8">8</a>].
What are the limitations of CSF analysis in CNS neoplasia?
CSF cytology has low sensitivity for detecting neoplastic cells because many CNS tumors do not shed cells into the CSF. Protein elevation and mononuclear pleocytosis are nonspecific findings in neoplastic disease [<a href="#ref-2">2</a>].