Multivariate methods in pharmaceutical applications

Abstract

This review covers material published within the field of pharmacy in the last five years. Articles concerning experimental design, optimization and applications of multivariate techniques have been published, from factorial designs to multivariate data analysis, and the combination of the two in multivariate design. The number of publications on this topic testifies to the good results obtained in the studies. Much of the published material highlights the usefulness of experimental design, with many articles dealing with optimization, where much effort is spent on getting useful results. Examples of multivariate data analysis are comparatively few, but these methods are gaining in use. The employment of multivariate techniques in different applications has been reviewed. The examples in this review represent just a few of the possible applications with different aims within pharmaceutical applications. A number of companies are using experimental design as a standard tool in preformulation and in combination with response surface modeling. The properties of e.g. a tablet can be optimized to fulfill a well-specified aim such as a specific release profile, hardness, disintegration time etc. However, none of the companies apply multivariate methods in all steps of the drug development process. As this is still very much a growing field, it is only a question of time before experimental design, optimization and multivariate data analysis are implemented throughout the entire formulation process, from performulation to multivariate process control.

Process analytical technology (PAT) for biopharmaceuticals

“…Jarka Glassey 1 , Krist V. Gernaey 2 , Christoph Clemens 3 , Torsten W. Schulz 3 , Rui Oliveira 4,5 , Gerald Striedner 6 , Carl-Fredrik Mandenius 7 goals [4,5]. In QbD the acceptable ranges of the CQAs and the CPPs are defined for the manufacturing process (Fig.…”

“…A detailed description of these techniques is beyond the scope of this contribution and can be found in a variety of textbooks and manuscripts (e.g. [6][7][8]). Their ability to reduce dimensionality by removing the redundancy and noise leads to the identificaton of salient features in the data.…”

Análise exploratória aplicada no estudo de medicamentos contendo piroxicam

“…Estes méto-dos têm possibilitado modelar propriedades químicas e físicas de sistemas simples ou complexos a partir de seus dados espectroscópicos, como no caso do estudo de formulações farmacêuticas (Gabrielsson, 2002;Pasquini, 2003;Yoon et al, 2004). Por outro lado, as ferramentas de Análise Multivariada permitem modelar sistemas onde estão sobrepostas informações espectrais ou cromatográficas de vários componentes através de várias regiões espectrais selecionadas para análise (Sabin, Banet, Ferrão, 2003;Sabin, Ferrão, Furtado, 2004).…”

“…Os dados são representados por uma matriz, onde as linhas representam as amostras e as colunas as variáveis (comprimentos de onda) (Lavine, 2002;Sabin, Banet, Ferrão, 2003). Pode-se empregar métodos de compressão da dimensionalidade como a Análise por Componentes Principais (ACP), ou simplesmente calcular a distância n-dimensional entre cada linha da matriz (amostra), empregando-se conceito de similaridade Gabrielsson, Lindberg, Lundstedt, 2002;Silva et al, 2002).…”

Desarrollo de un modelo SIMCA para la clasificación de kerosinas mediante el empleo de la espectroscopía infrarroja

“…16,17 El método SIMCA de reconocimiento de patrones (Soft Independent Modeling of Class Analogy) desde su introducción por Svante Wold en 1976 18 ha sido uno de los métodos de clasificación más utilizados en el control de calidad en las industrias alimenticia 19 y farmacéutica. 20 En la industria del petróleo se ha utilizado para la detección de adulteraciones en gasolinas. 10,13 Es un método supervisado de reconocimiento de patrones que se basa en el principio de analogía entre las muestras que pertenecen a una misma clase, y emplea para el cálculo de las distancias los scores determinados mediante análisis por componentes principales (PCA).…”