Critique of the determination of proteins in cerebrospinal fluid. Evaluation of the biuret method of Goa and the TCA-turbidimetric method of Meulemans.

‘]THE QUANTITATIVE DETERMINATION of total proteins in cerebrospinal fluid (CSF) has great clinical importance (1). Methods currently employed are not completely satisfactory, as is indicated by the frequent reports of modifications. Existing methods vary greatly both in accuracy and in complexity, with micro-Kjeldhal technics giving the most accurate results and being generally accepted as “ultimate” reference values (2-7). Kjeldhal procedures require as much as 15 ml. of sample (4), are difficult to perform (4), and are time consuming, as much as 2-3 hours being necessary for acid digestion (4, 8). Such procedures are impractical for routine use because hospital laboratories must provide analyses of CSF promptly. Many attempts have been made to develop a method as accurate and yet simpler and more practical than the micro-Kjeldhal determinations. The tyrosine-equivalence method of Johnston and Gibson, less involved than the micro-K jeidhal one but more so than certain other procedures, may also be employed as a “reference” since