Research paper Micro-CT in drug delivery Yiwei Wang a , David F. Wertheim b , Allan S. Jones c , Allan G.A. Coombes a, * a The University of Queensland, School of Pharmacy, Brisbane, Qld, Australia b Faculty of Computing, Information Systems and Mathematics, Kingston University, Kingston upon Thames, Surrey, UK c Australian Key Centre for Microscopy and Microanalysis, University of Sydney, NSW, Australia article info Article history: Received 12 January 2009 Accepted in revised form 14 May 2009 Available online 22 May 2009 Keywords: Micro-CT Drug delivery Matrix abstract Micro-computed tomography (micro-CT) has not to date been fully exploited in the area of controlled drug delivery despite its capability for providing detailed, 3-D images of morphology and the opportunity this presents for exploring the relationships between delivery device formulation, structure and perfor- mance. Micro-CT was used to characterize the internal structure of polycaprolactone (PCL) matrix-type devices incorporating soluble particulates (lactose Mw 342.30, gelatin Mw 20–25 kDa) as models of hydrophilic bioactives or pore-forming excipients. Micro-CT images confirmed that the lactose and gel- atin particles were uniformly dispersed throughout the PCL phase and that efficient delivery of 95– 100% of each species in 9 days involved transport from the matrix core. Quantitative analysis of micro- CT images provided values for matrix macroporosity, which were within 15% of the theoretical value and revealed uniform porosity throughout the samples. Total release of protein occurred in 9 days (PBS, 37 °C) from matrices containing a high protein load (44% w/w) and was independent of particle size. Measurements of equivalent pore diameter and frequency distribution identified a large population of sub-40 lm pores in each material, indicative of a high density of connecting channels between particles which facilitates protein transport through the matrices. Ó 2009 Elsevier B.V. All rights reserved. 1. Introduction Controlled drug release devices are designed to release bioac- tives at a defined rate, over a set time period to achieve therapeutic effects at local or systemic level. Implantable, topical and inser- table devices are often based on matrix-type designs in which drug particulates are distributed throughout a polymeric phase (Fig. 1). Such devices have been widely investigated for use in hormone replacement therapy, contraception, ocular drug delivery, treat- ment of cancer, control of bacterial infection [1–5] and more re- cently for production of tissue engineering scaffolds which incorporate growth factors [6]. The overall rate and pattern of drug release from matrix systems are determined by various factors including the physico-chemical properties of the drug and poly- mer, the pore structure of the matrix (pore size, percentage poros- ity, connectivity and tortuosity) (Fig. 1), drug diffusion through fluid-filled pores and channels (and/or the polymer itself), polymer dissolution, erosion or degradation. The kinetics of drug release from matrix-type devices incorporating dispersed drug in particu- late form (where release occurs through fluid-filled pores in the matrix) are frequently controlled by Fickian diffusion and conve- niently described by the Higuchi equation [7]. Q ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi tCsD e s ð2C  eCsÞ r where Q is the amount of drug released in time t, C is the initial drug concentration, Cs is the drug solubility in the release medium, D is the diffusion coefficient of drug molecules in the release medium, s is the tortuosity, e is the matrix porosity. Various techniques have been applied to characterize the poros- ity of materials including theoretical approaches, scanning electron microscopy (SEM), mercury porosimetry, gas pycnometry and adsorption. However, there are significant practical issues associ- ated with these measurements. Techniques such as mercury poros- imetry and gas flow porometry can be used to estimate pore size distributions but these typically differ by an order of magnitude due to differences in the underlying physics of the techniques [8]. Electron and other microscopies are extensively used to pro- duce images of porous materials but the challenge is again to quan- tify the dimensions of structural features such as pores and connecting channels having irregular shapes and sizes. Further- more, neither mercury porosimetry nor gas adsorption methods can account for closed pores, whilst mercury porosimetry only measures the distribution of constrictions in a pore network. Pore tortuosity (defined as the ratio of the actual path length through connected pores to the Euclidean distance (shortest linear distance) plays a key role in controlling drug delivery from matrix- type devices but is rarely quoted in the literature. Calculation of 0939-6411/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.ejpb.2009.05.008 * Corresponding author. The University of Queensland, School of Pharmacy, St. Lucia, Brisbane, Qld 4072, Australia. Tel.: +61 7 33451372; fax: +61 7 33651688. E-mail address: a.coombes@pharmacy.uq.edu.au (A.G.A. Coombes). European Journal of Pharmaceutics and Biopharmaceutics 74 (2010) 41–49 Contents lists available at ScienceDirect European Journal of Pharmaceutics and Biopharmaceutics journal homepage: www.elsevier.com/locate/ejpb