Comparison among different approaches for sampling cerebrospinal fluid in rats.

Two approaches for time-resolved sampling of cerebrospinal fluid (CSF) in rats were compared regarding performance, reproducibility, and extent of the inevitable trauma caused by the implantation of a sampling tube. Several parameters were checked to evaluate the injury: blood cell contamination of CSF; concentrations in CSF of the cytosolic proteins neuron-specific enolase (NSE) and S-100 (chiefly present in astrocytes); blood-brain barrier leakage of a dye-protein complex; viability of nervous tissue cells as assessed by dye exclusion; light and electron microscopy. In one sampling method, a tube was forced epidurally into the cisterna magna via a hole in the calvarium, consistently damaging the meninges and the nervous tissue. When using the alternative sampling method, the tube was instead affixed to the posterior atlanto-occipital membrane and connected with the cisterna magna via a hole in the membrane. Such a procedure caused negligible damage. Both techniques induced an inflammatory response. We advocate the use of the second approach, i.e., to sample CSF via a hole in the atlanto-occipital membrane, as the method of choice due to its high reproducibility. It is fairly rapid, and associated with a negligible injury.