Artificial neural network combined with principal component analysis for resolution of complex pharmaceutical formulations.

A chemometric approach based on the combined use of the principal component analysis (PCA) and artificial neural network (ANN) was developed for the multicomponent determination of caffeine (CAF), mepyramine (MEP), phenylpropanolamine (PPA) and pheniramine (PNA) in their pharmaceutical preparations without any chemical separation. The predictive ability of the ANN method was compared with the classical linear regression method Partial Least Squares 2 (PLS2). The UV spectral data between 220 and 300 nm of a training set of sixteen quaternary mixtures were processed by PCA to reduce the dimensions of input data and eliminate the noise coming from instrumentation. Several spectral ranges and different numbers of principal components (PCs) were tested to find the PCA-ANN and PLS2 models reaching the best determination results. A two layer ANN, using the first four PCs, was used with log-sigmoid transfer function in first hidden layer and linear transfer function in output layer. Standard error of prediction (SEP) was adopted to assess the predictive accuracy of the models when subjected to external validation. PCA-ANN showed better prediction ability in the determination of PPA and PNA in synthetic samples with added excipients and pharmaceutical formulations. Since both components are characterized by low absorptivity, the better performance of PCA-ANN was ascribed to the ability in considering all non-linear information from noise or interfering excipients.

Assessment of drug entrapment within liposomes using photophysical probes.

The photophysical and photochemical behavior of (R)-cinacalcet (CIN) and (S)-naproxen (NPX) entrapped within liposomes has been studied. For this purpose, liposome encapsulated drugs have been prepared through thin layer evaporation and characterized by transmission electron microscopy, cryoscopy scanning electron microscopy and dynamic light scattering. Steady state and time-resolved fluorescence experiments showed similar spectra, emission quantum yields, singlet energies and lifetimes for the selected drugs, outside and inside liposomes. By contrast, laser flash photolysis experiments revealed an important enhancement of the triplet lifetimes for entrapped drugs inside liposomes, indicating the spatial confinement existing in the microenvironment prevailing in these biomimetic entities. Thus, this photophysical property shows potential as a non-invasive, direct and valuable tool to monitor encapsulation of photoactive drugs and to probe the intraliposome environment. In addition, it provides a new quantitative indicator of the capability of liposomes to act as drug carriers.

Kinetic studies of nitrofurazone photodegradation by multivariate curve resolution applied to UV-spectral data.

This work aims at describing the kinetic model of nitrofurazone photodegradation by a novel chemometric technique, hybrid hard-soft multivariate curve resolution (HS-MCR). The study was applied to UV-spectral data from the photolysis of nitrofurazone solutions at different concentrations and exposed under varying illuminance power. The HS-MCR method was able to elucidate the kinetics of the photodegradation process and to determine the rate constants, and estimating at the same time the pure spectra of the degradation products. Exposure to light of the drug gave a first rapid isomerization to the syn-form that in turn underwent degradation furnishing a mixture of yellow-red products. The photodegradation process could be explained with a kinetic model based on three consecutive first-order reactions (A>B, B>C and C>D). These results were confirmed by application of the MCR procedure to the analysis of the data obtained from HPLC-DAD analysis of the nitrofurazone samples at different reaction times. The kinetic model was observed to be dependent on experimental conditions. The samples at higher concentrations showed rate constants lower than the diluted samples, whereas an increase of the rate of all degradation processes was observed when the light power also increased. This work shows the power of the hybrid hard- and soft-multivariate curve resolution as a method to deeply study degradation processes of photolabile drugs.

Prediction of photosensitivity of 1,4-dihydropyridine antihypertensives by quantitative structure-property relationship.

A quantitative structure-property relationships (QSPR) model, correlating the light sensitivity against theoretical molecular descriptors, was developed for a set of 1,4-dihydropyridine calcium channel antagonist drugs. These compounds are characterized by a high tendency to degradation when exposed to light, furnishing in the most of cases a related oxidation product from aromatization of the dihydropyridinic ring. Photodegradation was forced by exposing the drugs to a Xenon lamp, in accordance with the ICH international rules, and degradation kinetics was monitored by spectrophotometry. The photodegradation rates combined with a series of descriptors related to the chemical structures were computed by Partial Least Squares (PLS) multivariate analysis. An accurate selection of the variables, fitting at the best the PLS model, was performed. Two descriptors related to the substituent information on both the dihydropyridinic and benzenic rings and four molecular descriptors, were selected. The QSPR model was fully cross validated and then optimized with an external set of novel 1,4-dihydropyridine drugs, obtaining very satisfactory statistical results. The good agreement between predicted and measured photodegradation rate (R(2)=0.8727) demonstrated the high accuracy of the QSPR model in predicting the photosensitivity of the drugs belonging to this class. The model was finally proposed as an effective tool to design new congeneric molecules characterized by high photostability.

Determination of trapidil in human serum and urine by derivative UV spectrophotometry after selective solid-phase extraction.

A novel analytical technique able to determine the anti-ischemic drug trapidil in human serum and urine is proposed. In order to achieve satisfactory sensitivity and selectivity, an extraction procedure was required to isolate the drug from complex matrixes such as serum and urine. A solid-phase extraction procedure was investigated to both increase the analyte concentration and eliminate the interfering molecules present in large amounts in both matrixes. Optimization of the extraction step was realized by selecting a new polymeric sorbent based on a surface-modified styrene-divinylbenzene polymer which provided fast and efficient drug extraction. Drug quantification was performed by using the third-order derivative spectra of the SPE eluates. Absorbance specific signals at (3)D(335,316) and (3)D(316) nm for urine and serum, respectively, were demonstrated to be directly proportional to drug concentration and barely affected by residual matrix interferences. Under the optimized experimental conditions the calibration plots were linear over the concentration range 0.2-50 microg mL(-1). The method was validated by analysis of a series of spiked samples. Accuracy (recovery of 95 and 94% for serum and urine, respectively) and precision (RSD below 4%) were good.