Predictive, error-compensating kinetic method for enzymatic quantification of creatinine in serum

Here we describe an error-compensating kinetic-based method for the enzymatic quantification of creatinine in serum. The method, which has a large linear range and very low dependency on experimental variables that influence enzyme activity, is based on the use of creatinine amidohydrolase in a four-step coupled reaction sequence to generate a product that is monitored photometrically. We collected data for absorbance vs time during two to four half-lives of each reaction and fit a first-order model to the data to compute the total absorbance change that would be measured if the reaction were monitored to completion. Computed values of absorbance change agreed well with measured values and varied linearly with creatinine concentration in the sample throughout the range examined: 44 to 1326 mumol/L. A twofold change in enzyme activity in the final reaction mixture (3.3 to 6.6 kU/L) produced changes of only 12% and 4% for creatinine concentrations of 44 and 353 mumol/L, respectively. Results (y) for 39 serum samples that contained creatinine concentrations between 20 and 1800 mumol/L agreed well with liquid-chromatographic results (x), yielding linear least-squares statistics of y = (1.03 +/- 0.01) x + (5 +/- 5) mumol/L (r = 0.995, Sy.x = 37 mumol/L). We conclude that the predictive kinetic approach is a robust method for the quantification of creatinine in serum.