A Cerebrospinal Fluid Bank.

It has been found relatively simple to establish a cerebrospinal fluid bank. Spinal fluid removed at the time of pneumoencephalography may be preserved under sterile conditions by freezing and later used for replacement of cerebrospinal fluid lost at operation. The importance of the composition of solutions used to irrigate the brain or replace lost cerebrospinal fluid has been known for a long time. Forty years ago, Flexner and Amoss (1917) showed that intraspinal injection of normal saline, Ringer's or Locke's solutions increased the susceptibility of monkeys to polio virus while the reinjection of spinal fluid from other animals did not regularly do so and the reinjection of autogenous spinal fluid never did. Weed and Wegeforth (1919) reported that irrigation of the subarachnoid space in cats with a modified Ringer's solution produced no apparent ill-effects while irrigation with normal saline produced severe disturbances, convulsions, states of acute mania, and often death. Kasahara (1924) studied the effect of different solutions on the spinal subarachnoid space of rabbits using the cerebrospinal fluid cell count as the criterion. He found that 06 to 0-7% sodium chloride solutions were less irritating than other concentrations of saline. Ringer's solution was less irritating than 07% saline and human cerebrospinal fluid was the least irritating of any substances used. In 1934, Hartmann reported a method of preparing an artificial spinal fluid which Sachs (1945) used for replacement of spinal fluid lost from the ventricles at the time of operations for hydrocephalus. The preparation of this carefully buffered fluid required the preparation of two separate solutions which were separately sterilized and mixed just before using. Apparently the use of this fluid never became widespread, perhaps because of the technical difficulties of preparation. Elliott and Jasper (1949) studied the effects of irrigating fluids on the pial blood vessels of cats. They described the preparation of two sterile solutions for irrigation of the brain during operation and the replacement of fluid lost from the ventricles. One of these was a modified Ringer's solution, the other closely resembled Hartmann's solution but was more easily prepared. Lewis and Elliott (1950) later described the clinical use of this latter fluid, Solution B of Elliott and Jasper. It occurred to one of us (H.C.V.) that human cerebrospinal fluid removed at routine pneumoencephalography might be kept sterile and preserved by freezing. An ample supply of such fluid would then be available for the replacement of fluid lost from the ventricles at operation and for the actual irrigation of exposed brain tissue during operation. Initial studies were performed on specimens of human spinal fluid collected at the time of pneumoencephalography. These were frozen and stored for varying periods of time at -10°C. Chemical analyses of the spinal fluid were made at the time of collection and aliquots were then taken for analysis at monthly intervals during the period of storage. Sodium and potassium concentrations were determined with an internal-standard type flame photometer (Baird Associates), chloride by the method of Schales and Schales (1941), and pH with a Beckman pH meter. Total osmotic activity of the spinal fluid was determined by the measurement of the freezing point depression (Fiske Associates). Table 1 summarizes the results obtained. In each case the first figures represent the values obtained at the time of initial analysis. Interim determinations performed at four-week intervals did not differ significantly. It will be noted that initially the values for the pH are higher than those usually observed in normal blood serum. This undoubtedly occurred because no precautions were taken to prevent the loss of carbon dioxide from the spinal fluid. On freezing and thawing the fluid at the time of later analyses the value for the pH rose further, probably because of a similar mechanism with further loss of carbon dioxide.