Drug loaded microsphere elution is traditionally characterised using free-flowing methods comprising various geometric designs. Although informative for assessing burst elution under flow, these methods often neglect the in situ exchange kinetics of tissue contact (Fig 1). An in vitro method has been developed for real-time non-destructive fluorescent quantification of drug elution in a parenchymal tissue mimic for use in vitro to characterise diffusion rates, thereby potentially reducing the need for pre-clinical testing during formative device development. An alternative method involving physical sectioning of the tissue mimic was also employed for confirmation of drug diffusion rate, this approach could provide further supplementary information about doxorubicin release from Drug Eluting Bead’s (DEB’s) in a less transparent tissue-mimicking hydrogel, although this is not conducted in real-time.
Microspheres (radiopaque and non-radiopaque) were loaded with 37.5mg/mL Doxorubicin. The gel phantom matrix  (Fig 2) was composed of 0.5 %: 1 % alginate: agarose containing a single 1.6 mm ID circular channel set into a 40 cm circular bespoke cell contained with a semi-permeable Mylar® window. Once the gel was cast into the required cylindrical configuration the doxorubicin-loaded beads were filled into the lumen (shown in Fig 3). A custom SIOS detector with a working distance of 15-20 mm was employed to monitor drug diffusion from the beads. In a parallel experiment, 1mm thick sections were taken at predetermined time points to sample doxorubicin distribution in the hydrogel around the lumen and images were taken with a Olympus SZX16 fluorescence microscope fitted with SC50 camera (2.5x mag with GFPA filter) and processed with grey scale signal analysis for intensity.
(1). Drug diffusion by non-destructive real time laser confocal method
There was a non-uniform diffusion rate observed along the length of the gel (Fig 4), with statistically higher concentrations presented at the inlet channel of the bead column (p= <0.001, n= 3). A drug front with minimum 30 µM signal threshold, was detected 1141µm for non-radiopaque beads over the 2 hour period compared, with 858 µm for radiopaque beads (p= 0.0036, n= 3) (Fig 5). A zero order model was applied to both DEB types reporting a rate constant of 0.519 min-1 and 0.345 min-1 for non-radiopaque and radiopaque beads respectively, supporting a semi-infinite diffusion rate. The slower elution for radiopaque beads is aligned with novel flow-through dissolution methods [2, 3], and is likely due to tighter molecular packing resulting from the added radiopaque moiety slowing ion exchange with the gel matrix.
(2). Drug diffusion measurement by direct sectioning hydrogel
Figure 6 shows profiles of doxorubicin diffusion in the alginate-agarose gel measured by direct sectioning of gel samples containing DEB’s. It is observed that the diffusion of doxorubicin released from non-radiopaque beads were much faster compared to radiopaque beads with almost double the diffusion distance in 120 minutes. This is due to the high chemical potential of doxorubicin released from the less dense structure of non-radiopaque bead compared to the more compact radiopaque bead which contained iodinated radiopaque groups.
As a supplementary point, although the direct sectioning method is an invasive test method, it does provide consistent evidence of doxorubicin diffusion in tissue-mimicking hydrogel as the previous in situ confocal measurement. The tissue models also have potential for further refinement in terms of viscosity, ionic content and base geometry.
Conclusions: Two novel methods were used to monitor drug elution in a parenchymal tissue-mimic. The results demonstrated comparable trends to published radiopaque elution data in other models with intra-methodological alignment between section and real-time methods. Elution rate variations between microsphere types illustrate the sensitivity of the method for tissue diffusion rate determination. Tighter molecular packing in radiopaque beads results in a slower drug elution rate relative to the non-radiopaque beads tested in this experiment, potentially providing a longer therapeutic window in treatment.
References:  Bian, S., Seth, A., et al. (2017) Rev Sci Instrum. 88.(3): 034302
 Ashrafi, K., Y. Tang, et al. (2017) JCR.250: 36-47.
 Swaine, T., et al. (2016) Eur. J. of Pharm. Sci. 93: p. 351-359.