Multivariate analysis as a method to understand variability in a complex excipient, and its contribution to formulation performance. (26th November 2018)

• Record Type: Journal Article
• Authors: Tobyn, Mike; Ferreira, Ana Patricia; Lightfoot, Jane; Martin, Elaine B.; Ghimire, Manish; Vesey, Charles; Kasuboski-Freeman, Andrea; Rajabi-Siahboomi, Ali
• Is Part Of: Pharmaceutical development and technology. Volume 23:Number 10(2018)
• Page Start: 1146
• Page End: 1155
• Publication Date: 2018-11-26
• Subjects: Controlled release; multivariate analysis; formulation; ethylcellulose; Drug delivery systems -- Periodicals; Pharmaceutical technology -- Periodicals; Drugs -- Administration -- Research -- Periodicals; Drug Delivery Systems -- Periodicals; Pharmaceutical Preparations -- Periodicals; Technology, Pharmaceutical -- Periodicals; 615
• DOI: 10.1080/10837450.2018.1534862
• Languages: English
• ISSNs: 1083-7450

Abstract

Abstract: A key part of the Risk Assessment of excipients is to understand how raw material variability could (or does) contribute to differences in performance of the drug product. Here we demonstrate an approach which achieves the necessary understanding for a complex, functional, excipient. Multivariate analysis (MVA) of the certificates of analysis of an ethylcellulose aqueous dispersion (Surelease) formulation revealed low overall variability of the properties of the systems. Review of the scores plot to highlight batches manufactured using the same ethylcellulose raw material in the formulation, indicated that these batches tend to be more closely related than other randomly selected batches. This variability could result in potential differences in the quality of drug product lots made from these batches. Manufacture of a model drug product from Surelease batches coated using different lots of starting material revealed small differences in the release of a model drug, which could be detected by certain model dependent dissolution modeling techniques, but they were not observed when using model-independent techniques. This illustrates that the techniques are suitable for detecting and understanding excipient variability, but that, in this case, the product was still robust.