Doxorubicin eluting beads-2: methods for evaluating drug elution and in-vitro:in-vivo correlation.

ABSTRACT

DC Bead is a sulfonate-modified, PVA-based microspherical embolisation agent approved for the treatment of hypervascular tumours and arterio-venous malformations. The beads have previously been shown to actively sequester oppositely charged drugs, such as doxorubicin hydrochloride (dox) by an ion-exchange mechanism. In order to characterise the release kinetics and predict the in vivo behaviour of drug eluting beads (DEB), two elution methods were utilised. The first, an application of the USP dissolution method Type II - Apparatus, enables study of the complete elution of loaded DC Bead in less than 4 h, allowing relatively rapid comparison to be made between different products and formulations. Release data obtained using this method were fitted to first order kinetics (R (2) > 0.998) and the elution constants shown to increase with the total surface area of the beads exposed to the elution medium. Diffusion coefficients were calculated adopting the Fickian diffusion model, which predicted slow elution rates under physiological conditions. The second method involved the use of a T-Apparatus where the drug experiences an element of diffusion through a static environment. This method was developed to resemble the in vivo situation in embolisation procedures more closely. Slow release of dox from DC Bead with half-lives over 1,500 h were predicted for all size ranges using a slow release model. A strong linear relationship was found between the release data from T-Apparatus and pharmacokinetic data obtained from patients treated with DC Bead loaded with dox in transarterial chemoembolisation (TACE) procedures. These data indicated a Level A in vitro-in vivo correlation (IVIVC) for the first 24 h post embolisation. Both systems developed were automated and good reproducibility was obtained for all samples, demonstrating the usefulness of these elution techniques for product development and comparative testing.