Cerebrospinal fluid examination

Cerebrospinal fluid (CSF), also called liquor cerebrospinalis, is a transparent, colourless fluid (Fig. 1) contained within the subarachnoid space, bounded by two of the three membranes that protect the central nervous system (CNS): the arachnoid (externally) and the pia mater (internally). This space continues for a brief tract around the blood vessels that penetrate the parenchyma of the brain and the spinal cord. The other structures in which it is possible to find CSF are, in a caudo-rostral direction, the central canal of the spinal cord, the fourth ventricle, the mesencephalic aqueduct, the third ventricle and the lateral ventricles. (Figs. 2 and 3).

• Taking a sample of cerebrospinal fluid
• Examination of cerebrospinal fluid

Functions of cerebrospinal fluid – The CSF has numerous functions of which the foremost is to protect the CNS from impact against the bony walls in which it is enclosed. In fact, thanks to the CSF and the support given by the meninges, the effective weight of the buoyant CNS is less than its real weight without this system of amortisation. In fact, if the CNS were to be removed from the CSF-containing structure in which it is enclosed, it would undergo considerable changes to the point of rupturing through the effect of its own weight.

Other, no less important functions performed by the CSF are those of supplying nutrition to the CNS, regulating intracranial pressure and maintaining a more stable ionic environment than that of the plasma and, therefore, one more appropriate for the nervous system parenchyma. The CSF also has a metabolic role by acting as a method of transport of metabolites and nutrients between the brain and the blood.

Composition of the cerebrospinal fluid – The main component of the CSF is water, in which ions, nutrient substances and neurotransmitters are dissolved. Compared to plasma, the CSF contains lower concentrations of glucose and proteins, with albumin being the predominant protein. Despite their different compositions, CSF and plasma have the same osmolarity. The few cells physiologically present in CSF are mainly leukocytes (Table 1).

Table 1 -  Physiological characteristics of cerebrospinal fluid in the dog.

Production, flow and absorption – CSF is produced mainly by the choroid plexuses and, to a lesser extent, by the blood vessels of the pia mater and arachnoid. The choroid plexuses, located within the lateral, third and fourth ventricles, are villous formations, consisting of connective tissue, blood vessels, nerve endings and an epithelial covering (Fig. 4).

The CSF is produced by ultrafiltration of the blood plasma, by passage through the cuboidal epithelium of the choroid plexuses, and is modified by secretions from this same epithelium. The CSF circulates in the subarachnoid space and in the ventricular system by a flow which, exclusively for explanatory convenience, is described as starting from the lateral ventricles, before passing into the third ventricle through the interventricular foramina, after which it crosses the mesencephalic aqueduct to reach the fourth ventricle. During this journey, the choroid plexuses of the third and fourth ventricles add new fluid that they have produced, with the plexus of the fourth ventricle producing the major part. From here the CSF progresses towards the central canal of the spinal cord, although most passes through the lateral foramina of the fourth ventricle into the subarachnoid space. The CSF is absorbed in part by the veins suspended in the subarachnoid space by filaments that leave the pia mater and the arachnoid. Another site of absorption of the CSF, hotly debated in the past, is the arachnoid villi, prolongations of the subarachnoid space within the venous sinuses, located between the dura mater and the periosteum in the head and in the epidural space in the spinal cord.

Here the barrier between the blood and CSF is formed only of the arachnoid and the endothelium of the venous sinus (Fig. 5).  This absorption is the most important mechanism of controlling intracranial pressure and the amount of CSF. If, for some reason, such as damage to the arachnoid villi caused by meningitis, thes e mechanisms of disposing of CSF are lost, the CSF accumulates in the subarachnoid space.

TAKING A SAMPLE OF CEREBROSPINAL FLUID

As far as regards the lumbar access, the spinal tap is performed by introducing the needle dorsally into the intervertebral space between the fifth (L5) and sixth (L6) lumbar vertebrae, or between the sixth and seventh (L7) lumbar vertebrae.

The patient should be given a general anaesthetic independently of the sampling site. In order to avoid iatrogenic contamination as a result of introducing the needle into the subarachnoid space, it is important that the site of the sampling is shaven and prepared for surgery and that the operator wears sterile gloves. With regards to the size of the needle to use for CSF sampling, this differs according to the sampling site and the size of the animal: the needle for tapping the cisterna magna in small dogs is 22 gauge, 40 mm long with a stylet, whereas for large dogs the needle is 22 gauge and 75 mm long; the needle for tapping the lumbar cistern should be 75 or 90 mm long, depending on the size of the animal. Up to 1 ml of CSF for every 5 kg of body weight can be removed without causing problems, even if the amount needed for the laboratory investigations is much smaller. The CSF is collected into a sterile test-tube without anticoagulant.

CSF must not be sampled if the animal is suspected of having raised intracranial pressure. If CSF were to be withdrawn in this case, the drop in pressure would shift the brain and/or cerebellum caudally. Herniation of the cerebellum through the foramen magnum or of the cerebral cortex beneath the tentorium can compromise respiration to the point of causing death because of the compression of the breathing centres, or to lack of recovery of consciousness of the animal because of damage to the reticular activating system.

EXAMINATION OF THE CEREBROSPINAL FLUID

A full analysis of the CSF involves assessment of its physical parameters (colour and turbidity), cytological analysis (number and type of cells) and biochemical analysis, useful in particular for determining the protein concentration. For routine analyses, 1 ml of CSF is usually sufficient.

Physical analysis: this is conducted by observing the CSF contained in a transparenttest-tube against a white background. In a normal subject the CSF is colourless, transparent and limpid. In some cases the colour may be different. For example, the CSF may be a reddish colour because of the presence of red blood cells. If this colour disappears when the test-tube is centrifuged it means that the bleeding occurred at the time of sampling the CSF. In pathological situations, a reddish colour may be due to a recent haemorrhage in the CNS, since red blood cells break down quickly, giving, in the case of older haemorrhages, a yellowish (xanthochromic) rather than reddish colour to the CSF. The yellowish colour is due to the presence of free bilirubin in the CSF, found after haemorrhage or jaundice (Fig. 6). The CSF becomes turbid when it contains a large quantity of protein and/or cells (>500 white blood cells/µl).

Cytological examination: this consists of quantitative and qualitative analyses to determine, respectively, the number and types of cells present in the sample of CSF.

Quantitative analysis (cell count); this can be performed using a cell counting chamber, such as a Neubauer chamber or a modified Fuchs-Rosenthal chamber (Fig. 7). This latter is formed of a glass support which has two grids, each of 3.2 mm_³, subdivided into 16 squares. In order to prepare the sample, 90 µl of CSF are placed in a cuvette together with 10 µl of fuchsin dye. Five minutes later, a small amount of the sample prepared in this way is made to move by the capillary effect between the Fuchs-Rosenthal chamber and glass coverslip, placed over the drop of CSF. With the help of an optical microscope, at 100x magnification, the operator counts the cells in one of the grids in the chamber, corresponding to 3 µl of CSF. The number of cells contained in 1 µl di CSF is therefore obtained by dividing the number of cells counted by three (Video 2).  Some operators carry out a direct reading of the CSF, placing a drop of the fluid in the Fuchs-Rosenthal chamber and making an immediate count (Fig. 8).

There are usually very few cells in the CSF of a healthy subject: generally the cells that are present are leukocytes and histiocytes. A cell count of up to 5 cells per microlitre is considered normal.

Quantitative cytological changes can consist of an increase in the number of cells, defined pleocytosis, which ranges from just a few cells to several hundred cells. Pleocytosis is classified as mild (<50 cells/µl), intermediate or moderate (50-200 cells/µl) and severe (>200 cells/µl). An increase in the number of cells in the CSF is generally associated with inflammatory processes of the CNS and the greater the involvement of the meninges or ependymal cells, the greater the increase in the cell count. In contrast, if the inflammatory disorder is confined to the nervous system parenchyma, there may be only a mild pleocytosis (<50 cells/µl), regardless of the severity of the disease, such as in distemper or rabies. Examples of severe pleocytosis are the increases in cell count occurring in steroid-responsive meningitis-arteritis and in granulomatous meningo-encephalomyelitis, conditions in which the cell count in the CSF may exceed 500 cells/µl.

The increase in the cellular component may also be due to contamination of the specimen by blood during the sampling procedure. When the contamination is modest, the leukocyte count is corrected by subtracting one white blood cell for every 500 red blood cells.

Qualitative analysis. Specimens of CSF  must be concentrated in order to obtain a sufficient number of cells to enable the cytological evaluation. In fact, a direct smear on a glass slide is diagnostically unhelpful because there are too few cells and not enough proteins to provide an adequate support for the cells, thus favouring phenomena of cellular distortion during the process of preparing the smear.

The different techniques for concentrating CSF are sedimentation and cytocentrifugation. Sedimentation has been shown to be very useful in outpatient practices because the qualitative and quantitative recovery of the cells is good, without being too expensive. Cytocentrifugation, which enables a better quality cytological examination, requires a cytocentrifuge in which the sample of CSF (50-100 µl) is centrifuged at 1000-1500 rpm for 5-10 minutes. In this latter case the instrument required is obviously much more expensive (Fig. 9).

Bacteria can be found in the CSF of animals with bacterial meningitis, albeit rarely: more commonly any bacteria found in the CSF are contaminants, even if they appear phagocytosed by neutrophils on the glass slide. Cryptococcus neoformans, a round or oval fungus has a diameter from 5 to 15 μm and a semitransparent refractive capsule, which appears characteristically colourless after staining with India ink (one drop of fresh CSF sediment plus one drop of India ink).

As far as concerns neoplasms of the CNS, cytological analysis of the CSF is of only limited importance for diagnostic purposes. This is because: (i) a tumour will not exfoliate cells into the CSF unless it is in communication with the subarachnoid space; (ii) benign primary tumours of the CNS exfoliate cells that are indistinguishable from normal cells; and (iii) cytological findings alone are only rarely sufficient to make a definitive diagnosis, given the often minimal atypia of the tumour cells. Examination of the CSF can, however, provide the diagnosis of lymphosarcoma, especially when this involves the meninges, because lymphomatous cells are easily distinguishable from normal inflammatory cells: the features differentiating them are that the former have bulky nuclei with irregular margins, large size, scarce basophilic cytoplasm, frequent mitoses and atypias (Fig. 10).

Pleocytosiscan be distinguished into neutrophilic, if there is a predominance of polymorphonuclear neutrophils, lymphocytic, if most of the cells are lymphocytes, eosinophilic, if most of the cells are polymorphonucleated eosinophils and mixed if there is an increase in various nucleated cells (lymphocytes, monocytes, macrophages, neutrophils, some eosinophils and plasma cells) (Fig. 11).

In the dog, the most common cause of a neutrophilic pleocytosis is steroid-responsive meningitis-arteritis, particularly in the acute phase of the disease: analysis of the CSF shows a dramatic increase in cellular production (often up to 1000-2000 cells/µl or more), which are almost all (80-100%) non-degenerated neutrophils. Bacterial meningitis and meningo-encephalitis also cause the same sort of pleocytosis. In these cases the CSF shows severe pleocytosis (>1000 cells/µl) with a predominance of degenerated neutrophils. The signs of degeneration involve the nucleus and consist of swelling, hyalinisation, karyolysis, and karyorrhexis, while bacteria may not necessarily be present in the sample. A similar picture with a prevalence of neutrophils can be seen in some neoplasms, particularly meningiomas; the cause may be tumour-associated necrosis.

Lymphocytic pleocytosis is associated with viral encephalitis. Examples of viral encephalitis in the dog are distemper, rabies and “central European tick-borne encephalitis” caused, respectively, by viruses of the genera Morbillivirus, Rhabdovirus and Flavivirus. Lymphocytic pleocytosis may also occur in necrotising encephalitis, with the CSF containing 200-500 cells/ml, 80-100% of which are lymphocytes.

In rare cares there may be a predominantly eosinophilic pleocytosis, which has been described in cases of steroid-responsive meningitis, in some tumours (for example, histiocytoma), infections by Toxoplasma gondii, Neospora caninum, and Cryptococcus neoformans (even though neutrophilic pleocytosis is more common), as well as in a rare form of idiopathic meningoencephalitis called eosinophilic meningoencephalitis  (Fig. 12)

A commonly found picture at the time of analysis of CSF is pleocytosis with a mixed population of lymphocytes, monocytes, macrophages, neutrophils, a few eosinophils and plasma cells. This picture has been described in many cases of meningo-encephalitis of unknown aetiology (Fig. 13), in toxoplasmosis, neosporosis, mycoses of the CNS (cryptococcosis, blastomycosis, aspergillosis) and in chronic forms of steroid-responsive meningitis.

Table 2 presents the CNS diseases that may be associated with different types of pleocytosis (Table 2).

Table 2: Cytological changes of CSF and differential diagnoses of CNS diseases. Classification of pleocytosis: mild, <50 cells/µl; moderate, 50-200 cells/µl; severe, >200 cells/µl. SRMA, steroid-responsive meningitis-arteritis; GME, granulomatous meningo-encephalitis. (Modified from Chrisman, 1992).

Biochemical analysis – Eighty percent of the proteins of the CSF are derived from the plasma, by ultrafiltration or active transport. The movement of these proteins from the blood to the CSF is affected by their size, electrical charge, concentration in the plasma and condition of the blood-brain barrier. The protein content also differs depending on where the sample of CSF is taken from (it is normally less than 30 mg/dl when taken from the cisterna magna and less than 45 mg/dl when taken from the lumbar cistern).

The total protein concentration can be assessed in various ways, some of which are easily reproducible in outpatient practice. Urine dipsticks can be used: these are more sensitive to albumin and demonstrate its presence by a change in colour of a pH indicator.

Another method is the Pandy test, which is performed by mixing a few drops of CSF with Pandy’s reagent (10 mg of carbolic acid crystals dissolved in 100 ml of distilled water). The globulins in the CSF are selectively precipitated producing a cloudy turbidity (Fig. 14). Another option, for those veterinarians without the possibility of carrying out the preceding tests, is to send the sample of CSF to a specialised laboratory which can quantify the protein content by a turbidimetric method (such as the pyrogallol red test), a colorimetric test (Lowry’s method) or Coomassie’s brilliant blue G-250 method (Bio-Rad).

An increase in the protein content is a generic indicator of a CNS disease and suggests either damage to the blood-brain barrier or an increase in the intrathecal production of immunoglobulins. For example, in inflammatory disorders the increase in total proteins occurs because of destruction of the tight junctions between the endothelial cells of the meningeal vessels and the nervous system parenchyma, making it possible for molecules, such as albumin, to cross the blood-brain barrier. In normal conditions, albumin, with its molecular weight of 65000 Dalton and above, would not be able to cross the blood-brain barrier. The increase in proteins is partly due to the breakdown of cells accompanying the inflammation. The closer the site of inflammation is to the meninges, or if the meninges are directly involved, the more likely it is that there will be a change in the protein content of the CSF, whereas this is less likely if the inflammation is confined to deep areas of the parenchyma.

IgG are considered the predominant class of immunoglobulins in the CSF and may originate from the serum, reaching the CSF through a damaged blood-brain barrier; alternatively, they may be of intrathecal origin, being synthesised locally.

A high concentration of protein in the CSF is often accompanied by a pleocytosis, but may also be found in the absence of an increase in the cell count: in this latter situation, the condition is named “albuminocytological dissociation”. An increase in total protein content together with a normal cell count can occur in non-suppurative encephalomyelitis due to distemper virus, in which the increase in the concentration of total proteins seems to be due to intrathecal production of immunoglobulins. A similar albuminocytological dissociation can occur in neoplasms, as a result of tissue necrosis, alterations in the blood-brain barrier or local production of immunoglobulins induced by the tumour itself.

CSF can also be used for serological analyses, for a search for antibodies, such as those against distemper virus, against some protozoa, including Toxoplasma or Neospora, or against some Rickettsia spp. Antibody titres are not always easy to interpret: for example, false positive results are possible if the animal has been vaccinated against a given antigen. In such situations it is useful to assay the IgM, as well as the IgG, and, possibly, repeat the examination a few times over the period of several weeks to detect any increases in antibody titre, which would indicate an ongoing immune response. Molecular biology methods, such as polymerase chain reaction analysis, can detect small amounts of foreign nucleic acids, for example, those from distemper virus.

Suggested readings

1. Bailey C.S., Higgins R.J.: “ Characteristic of cerebrospinal fluid associated with canine granulomatous meningoencephalomyelitis: a retrospective study.” Journal of American Veterinary Medicine Association, 1986; 188:418-421
2. Bailey C.S., Higgins R.J.: Characteristic of cisternal cerebrospinal fluid associated with primary brain tumors in the dog: A retrospective study. Journal of American Medical Association, 1986; 188 (4): 414-417.
3. Bernardini M: “Esami collaterali”  IN: Neurologia del cane e del gatto –seconda edizione. Poletto Editore, Padova – 2010, pp 85-94.
4. Bush W.W., Barr C.S., Darrin E.W., Shofer F.S., Vite C.H., Steinberg S.A.: “Results of cerebrospinal fluid analysis, neurologic examination, and age at the onset of seizures as predictors for results of magnetic resonance imaging of the brain in dogs examined because of seizures: 115 cases (1992- 2000)”. Journal of  American Veterinary Medical Association,2002, 220 (6): 781-790. 
5. Chrisman C.L.: “Cerebrospinal fluid analisys”. In: Moore MP (ed.), Disease of the spine, WB Saunders, Philadelphia PA:Veterinary Clinics of North America: Small Animal Practice, 1992; 22: 781-810.
6. DeLahunta A, Glass E: “Cerebrospinal Fluid and Hydrocephalus” In: “Veterinary Neuroanatomy andf Clinical Neurology” – 3^rd Ed, saunders, Philadelphia - 2008 pp54-76
7. Jenkins T.W.: “Meningi e liquido cefalorachidiano”. In: Neuroanatomia Funzionale Dei Mammiferi, Bologna, Calderini, 1989; 66-73.
8. King A.S.: “Meninges and cerebrospinal fluid”. In: Physiological and clinical Anatomy of the Domestic Mammals , Oxford, Oxford Science Ltd, 1987; 13-23.
9. Munana K.R.: “Encephalitis and meningitis”. In: Veterinary Clinics of North America: Small Animal Practice, 1996; 26: 857-874.
10. Radaelli T., Platt S.: “Bacterial Meningoencephalomyelitis in Dogs: A Retrospective Study of 23 Cases (1990-1999)”. Journal of Veterinary Internal Medicine, 2002; 16: 159-163.
11. Rusbridge C.: “Collection and interpretation of cerebrospinal fluid in cats and dogs”. In practice, 1997; 19 (6): 322-324, 327-331.
12. SorjonenD.C., Warren J.N., Schultz R.D.: “Qualitative and quantitative determination of albumin, IgG, IgM, IgA in normal cerebrospinal fluid of dogs”. Journal of the American Animal Hospital Association, 1981, 17: 833-839.
13. SorjonenD.C.: “Total protein, albumin quota, and electrophoretic patterns in cerebrospinal fluid of dogs with central nervous system disorders”. American Journal of Veterinary Research, 1987, 48:301.
14. Tipold A, Stein VM: “Inflammatory diseases of the spine in small animals”.  Vet Clin North Am Small Anim Pract. 2010 Sep;40(5):871-9.
15. Thomson C.E., Kornegay J.N., Stevens J.B.: “Analysis of cerebrospinal fluid from the cerebellomedullary and lumbar cistern of dog with focal neurologic disease: 145 cases (1985-1987)”. Journal of American Veterinary Medical Association, 1990; 196 (11): 1841-1844.
16. Tipold A., Jaggy A.: “Steroid responsive meningitis-arteritis in dogs: Long-term study of 32 cases”. Journal of Small Animal Practice, 1994; 34: 311-316.
17. Tipold A., Vandevelde M., Jaggy A.: “Neurological manifestations of canine distemper virus infection”. Journal of  Small Animal Practice, 1992; 33: 466-470.
18. Tipold A.: “Cerebrospinal Fluid”. In: Clinical Neurology in Small Animals- Localization. Diagnosis and Treatment. 2003, K.G. Braund, International Veterinary Information Service (www.ivis.org). Ithaca, New York, USA.
19. Tipold A.: “Diagnosis of Inflammatory and Infectious Disease of the Central Nervous System in Dogs: A Retrospective Study”. Journal of Veterinary Internal Medicine, 1995; 9: 304-314.