CSF analysis is an important diagnostic tool in the investigation of neurological patients, but there are limitations to it. The cell count and protein level of the CSF can be thought of as a CNS analogue of the CBC and serum protein level. Abnormalities in the colour, cellularity and protein level of the CSF are strongly indicative of neurological disease and often contribute to the diagnosis, but are non-specific and only occasionally provide a definitive diagnosis by themselves. CNS diseases also do not always cause alterations in the CSF – abnormalities depend on the location and extent of the lesion. Prior treatment with corticosteroids may interfere with CSF and reduce the white blood cell count and percentage of neutrophils. CSF should ideally be processed (or preserved) within 30 minutes after collection because the cells deteriorate rapidly due to little protein content.

CSF collection should be performed prior to myelography, since contrast agents may induce meningeal irritation and change the character of the fluid by producing an inflammatory response. Changes can influence CSF analysis for at least three to five days post myelogram.

Typically, a total cell count, differential cell count and a protein level are determined in the CSF. A glucose level is occasionally obtained and is normally 60-80% of the blood glucose level. When an infectious disease is suspected, culture, serology or PCR can be performed. In case of lymphoma, flow cytometry or PARR can be used to confirm the diagnosis.

Colour and clarity

Normal CSF is transparent, colourless and of water consistency.

A haemorrhage in the CSF, occurring a minimum of 10 hours before collection may result in xanthochromia – yellow-orange staining, caused by bilirubin from metabolized erythrocytes. This discolouration can persist up to 2-4 weeks post haemorrhage into the subarachnoid space but is usually resolved in 4-8 days. Other causes of xanthochromia can be severe icterus or markedly elevated CSF protein levels.

Blood contamination resulting in red coloured CSF may be iatrogenic or due to haemorrhage in the subarachnoid space. To simply differentiate between iatrogenic and pathological haemorrhage, a part of the sample can be centrifuged. A xanthochromic supernatant is indicative of pathologic haemorrhage, but a colourless one suggests iatrogenic contamination. Pathological haemorrhage can also be suspected when erythrophagia, haemosiderophages or haematoidin crystals are observed during microscopic examination.

Cloudy CSF can be due to markedly elevated number of cells (> 500 WBCs/µl) or increased protein levels, which will also cause the fluid to be more viscous.

Cell count

Normal WBC count in CSF is < 5 cells/µl. In differential count, mononuclear cells should predominate (60-70% of it being small lymphocytes and 30-40% monocytes) and only occasional mature neutrophils are found (< 1%, excluding blood contamination). Elevation of nucleated cells in the CSF is called pleocytosis.

Greater number of leucocytes are expected with diseases that involve meninges or ependymal cells, but that does not correlate with the disease severity or prognosis. Deep parenchymal, extradural and non-exfoliative lesions cause minimal or no change in the CSF or only increase protein levels.

Mononuclear (mixed) cell pleocytosis is an increased leucocyte count with a predominance of lymphocytes and macrophages. It is usually associated with granulomatous meningoencephalitis (GME) in dogs. Other causes of mixed cell pleocytosis are fungal, protozoal and rickettsial infections. The necrotizing meningoencephalitis (Pug/Maltese, Yorkshire Terrier encephalitis) usually cause primarily lymphocytic pleocytosis. This type of pleocytosis can be also seen in viral meningitis (rabies, canine distemper) and lymphoma in the CNS.

A neutrophilic pleocytosis is often associated with bacterial infections or steroid-responsive meningitis. Less commonly may neutrophils predominate in some viral (acute canine distemper, FIP) or fungal infections, meningiomas and fibrocartilaginous embolic myelopathy. In cases of infectious diseases, the neutrophils are more likely to be degenerated than in non-infectious disorders. Culture of CSF should be considered in animals with neutrophilic pleocytosis, especially if the animal is showing other signs suggestive of bacterial meningitis – pyrexia, peripheral neutrophilia and degenerated neutrophils in the CSF.

Eosinophilic pleocytosis is rare and can be associated with aberrant parasite migration in the CNS, protozoal, cryptococcal and protothecal infections, rabies or canine distemper, or rare eosinophilic meningoencephalitis (steroid-responsive).

Normal CSF should not contain erythrocytes, but a low number is commonly seen due to iatrogenic blood contamination. Blood contamination can interfere with the interpretation of WBC count and a correction formula can be used: every 500 RBCs/µl may account for 1 WBC/µl in a dog and every 100 RBCs/µl for 1 WBC/µl in a cat; but sometimes even erythrocytes as high as 15.000/µl do not significantly affect the leucocyte count.

Protein level

Normal protein concentration is higher in the lumbar region and generally has values < 45 mg/dl, compared to cisternal tap which contains < 25 mg/dl of protein. As the protein concentration in the CSF is much lower than in other body fluids, it cannot be measured by refractometer. The effect of haemorrhage on protein level is usually low – approximately 1200 RBCs/µl are needed to increase the protein concentration by 1 mg/dl.

A normal cell count with an elevated protein level is called albuminocytologic dissociation. It occurs with diseases that alter the blood-brain barrier and allow protein from the circulation to enter the CNS, increase the production of protein within the CNS or obstruct the flow of fluid in the CNS (extradural compressive lesion as IVDD, intramedullary mass, trauma, vasculitis, ischaemic CNS necrosis, degenerative myelopathy …).


Dewey C. W., Da Costa R. C.: Practical guide to Canine and Feline Neurology, 3rd Ed. Wiley Blackwell, 2016

Platt S. R., Olby N. J.: BSAVA Manual of Canine and Feline Neurology, 3rd Ed. BSAVA, 2004