ic sites are involved. However, these still remain open conjectures.
A brief description of a useful graphical technique to perform the calculations might also be of interest. The equation for total concentration can be rearranged to give:
This expression is a straight line; if it is plotted on the same graph as x = xfC), the intersection provides the desired value of free concentration corresponding to a (given) total concentration of drug present (see Fig. ). Shen and Gibaldi used an iterative computer routine to determine the solution of the nonlinear algebraic equation, but the simple graphical technique often provides answers of adequate accuracy, or they could be used as starting values for an iterative computer solution.
To conclude, the questions raised by Shen and Gibaldi concerning the effective protein concentration concept are valid and were the basis of our choice of the term "effective." We did not feel that the data we used (1954 vintage) were sufficiently precise, nor was there a wide enough range of information, to be worth more detailed analysis. We had hoped that more extensive and detailed investigations would be done to test the conjectures. It does seem, however, that good simulations and predictions of tissue drug levels are possible, both from our own work and that of others during the intervening several years. As discussed by Shen and Gibaldi, the alternate method of utilizing a tissue-plasma distribution ratio, R(c) varying with concentration for nonlinear binding and estimated from in uivo data, is probably the best approach in light of our present knowledge of binding.
However, this approach has limitations: (a) a fairly large amount of in uivo biopsy data is required, since predictions from in uitro binding isotherm data are presumably not used; and (b)
interesting effects, such as binding saturation and/or competitive binding of two drugs, cannot be predicted. Therefore, it is felt that the effective protein concentration concept has valuable and unique capabilities for pharmacokinetic simulations, and it is hoped that further development of it, along with fundamental binding studies with physiological protein concentrations, will lead to a more firm quantitative basis for handling drug binding.