Novel univariate and multivariate regression methods for the simultaneous analysis of ternary mixture of 1, 4-benzodiazepines in dosage forms and human urine

摘要

In this paper, novel univariate and multivariate regression model updating methods were developed and validated for the simultaneous determination of ternary mixture of bromazepam, diazepam and clonazepam. The univariate methods used are the mean centering of ratio spectrophotometric method and iso-absorptive point coupled with ratio-subtraction and ratio difference spectrophotometric method. Linearity, ranges, precision and accuracy of these methods were determined. The multivariate methods adopted are the principal component regression and partial least-squares. These models were estimated using nineteen mixtures as calibration set and six mixtures as validation set. The minimum root square error of predication were 0.1215 and 0.0568 for BMZ, 0.0598 and 0.0712 for DIZ and 0.0867 and 0.0753 for CNP by PLS and PCR respectively. PCR and PLS methods were successfully applied for the analysis of BMZ, DIZ and CNP in their dosage forms. Content uniformity testing of the studied pharmaceutical tablets by PLS and PCR was also determined. MCR, IPRSRD and PLS model updating enabled the determination of BMZ, DIZ and CNP in spiked human urine. Moreover, PLS model updating was applied to drug-dissolution rate testing of BMZ and DIZ in their commercial tablets. The obtained results were statistically compared with those obtained by reported methods giving a conclusion that there is no significant difference regarding accuracy and precision.