a patient is in an intensive care facility showing severe focal or diffuse brain injury, then consideration might be given to inserting a separate subarachnoid or subdural cranial catheter at the time of anesthesia for sampling CSF from the AO site to allow the assessment of intracranial pressure and the derivation of an estimate of cerebral perfusion pressure.27,28 This can be of use in determining the effectiveness of ongoing management of such cases. Patient‐side observations and tests are extremely important at this stage. For example, if initial CSF is pink but clears, iatrogenic hemorrhage is likely the cause. With a sick neonatal patient, cloudy CSF with a low glucose content detected on urine test strips likely indicates suppurative meningitis. Particularly during lumbosacral sampling, small particles can often be seen to swirl in the aspirated sample. These are usually determined to be epidural fat and small aggregates of collagenous tissue introduced into the sample at the time of collection, likely reasons for falsely elevated CSF‐CK activity. When CSF analysis cannot proceed immediately following collection, the addition of a small, quantified volume (~10%) of autologous serum to an aliquot of the sample may help support cells for better cytologic harvesting, at least for 24 h.
Of note is that estimates of protein content in CSF may be substantially higher, at least when tested by the pyrogallol red method, when CSF is collected into commonly used red vacutainer tubes containing clot activator.(29)
The results of CSF analysis can reflect diseases in the brain and spinal cord just as a hemogram detects many systemic disease states. Thus, CSF analysis is one of the most helpful aids in determining the nature of current and progressive CNS lesions. Normal CSF is clear and colorless with a refractive index of 1.3347–1.3350. In addition, it contains no erythrocytes and usually less than 6 × 106/L (=6/μL) small mononuclear cells. Protein content is usually between 0.5 and 1.0 g/L (=50–100 mg/dL) in the horse with an absolute range of 0.1–1.2 g/L depending on the laboratory and can be up to 1.8 g/L in normal neonatal foals. Ruminants and pigs usually have CSF with a protein content of less than 0.75 g/L.18 Specific electrophoretic protein fractions in CSF can be assayed by numerous methods,17 but the utility of these results in changing the clinical course of cases is doubtful at present. Neonates of all species may have slightly xanthochromic CSF.
To help account for the blood contamination of CSF samples and to help determine the local versus systemic source of measured CSF immunoglobulin content, the determination of the albumin quotient and the IgG index, respectively, can be of some use.30,31 Additionally, on the assumption that all red blood cells are from the circulation and not the result of previous subarachnoid bleeding, arithmetical corrections can be made for all measured components of CSF. Such corrections for routinely measured CSF components have not proven to consistently improve the utility of absolute measurements of amounts of CSF constituents.32–34 Repeat thecal puncture from the lumbosacral or from the atlanto‐axial space 2 weeks apart has been shown not to impact results of CSF analysis.35
Figure 3.7 Collection of CSF from the atlanto‐occipital cistern in a standing sedated horse. After the ears have been secured in a forward position and the area clipped and surgically prepared, a 3.5inch (8‐10 cm) 18g spinal needle is introduced and advanced as described for the procedure in recumbent anesthetized horses (A). When a distinct change in resistance (“pop”) is felt, the stylet is removed, and CSF is gently aspirated into a syringe attached to extension tubing (B). The needle may be advanced under ultrasound guidance as described by Depecker et al.20
Source: Images courtesy of Carol Clark, Peterson Smith Equine Hospital & Complete Care, Ocala, FL, USA.
Many constitutive and induced enzymes have activities expressed in CSF. The cytosolic BB‐CK isoenzyme almost certainly is quite specific for neuronal tissue damage, but its measurement is somewhat complex and the interpretation of results of its content in CSF in disease states is not extremely specific. Total CK activity in CSF may well vary with disease states,36 and the utility of such assays in general disease states has been questioned.37 In specific neuromitochondrial disorders, lactate and pyruvate measurements may well be of diagnostic use.
Assays are available for proteins in and antibodies to infectious agents in CSF.38,39 These tests are probably best for the subacute to chronic infections with particular viruses and protozoa. Indeed, testing for the presence of antibodies to Sarcocystis and Neospora surface antigens in CSF and blood has become popular and is very useful to assist in the diagnosis of equine protozoal myeloencephalitis,40,41 especially when Sarcocystis neurona vaccines were used.38 In particular, the ratios of specific antibody in serum to CSF are used to reveal locally produced (i.e., intrathecal) antibodies indicative of CNS infection.42 Unfortunately, while this antibody ratio technique is highly specific and sensitive for EPM diagnosis, it is not accurate for the diagnosis of equine neuroborreliosis.43 As with many of such tests, a negative result with S. neurona immunotesting using CSF (and serum) is very useful to rule out the disease in a case with clear clinical signs in an area where the background frequency of exposure is low. However, debate arises as to interpretation of marginal test results when clinical signs are equivocal and there is a high exposure rate in the sample population. This is particularly true for a negative result, as the negative predictive value declines with increasing disease (exposure) prevalence, and for EPM the consequences of a false‐negative result (failure to treat) are more serious than a false‐positive result (unnecessary treatment). Again, with a high rate of false‐positive and false‐negative results occurring with borreliosis testing, there can be no reliance on serology for the confirmation of that disease.
One aspect of CSF chemistry that may well become useful in future is the analysis of levels of neurotransmitters, neurohormonal metabolic products, and antibodies directed against constitutive proteins to help categorize some disease states.44–49 This will most likely relate to those diseases that have no recognized morbid neuropathologic basis, such as shivers, headshaking, narcolepsy, cataplexy, self‐mutilation, rage syndrome, acquired tremor syndromes, spastic syndrome, and spastic paresis, although some progress has recently been made on the search for a putative genetic component to the latter two syndromes.50
Figure 3.8 Ultrasound‐guided collection of CSF from between C1 and C2 in a standing sedated horse (A and B). In C, the spinal cord (SC) and dura mater (green arrows) are imaged using a transversely oriented ultrasound transducer. A 3.5inch (8‐10 cm) 18 g styletted needle is introduced from below the transducer and advanced into the subarachnoid space, then the stylet is removed and CSF is aspirated into a syringe attached to extension tubing. The dashed orange line in C indicates a typical dorsolateral to ventromedial path of needle placement. The “Z” symbol at the top marks the dorsal aspect of the ultrasound image.
Source: Images courtesy of Sally DeNotta, University of Florida, USA.
With most CNS malformations, the CSF analysis will be normal, although those anomalies that result in tethering of neural tissues such as complex vertebral deformations can result in progressive disease with ongoing traction injury to CNS neural tissues during growth, thus potentially resulting in CSF changes consistent with trauma. If a malformation of the calvaria or vertebral column damages underlying nervous tissue, the CSF may reflect compression with evidence of subtle hemorrhage.
Infectious diseases can result in CSF pleocytosis and the elevation of protein content. The cell type present varies considerably, but generally neutrophils predominate with bacterial diseases and small mononuclear cells with viral diseases. Notable exceptions to this are high neutrophil numbers with Eastern equine encephalitis and high