Implementing 3D Printing for the Long-Term Release of Biomacromolecules

The advancement of new biotherapeutics demands appropriate strategies that allow effective delivery to the body for the treatment of chronic diseases, which is of growing importance globally. This includes the development of implantable delivery systems with enhanced control over release at an intended site with the desired dosing. Additive manufacturing (AM), often referred to as three-dimensional (3D) printing, has been identified due to the potential advantages. This includes lower-cost initial formulation trials, geometric control and spatial location of bioactives to better modulate delivery in comparison to common manufacturing routes e.g. tabletting. The use of inkjet printing (IJP) systems has well-documented success for producing delivery devices that allow release of small molecules (Da), but this is less reported regarding delivery of larger bioactives (kDa). This prompted the initial exploration in this thesis into IJP using poly(ethylene glycol) diacrylate (PEGDA) based formulations to assess encapsulation and delivery of biomacromolecules from 3D structures. Challenges encountered with this approach included increases in viscosity with increases in protein loading as well as precipitation during droplet ejection, residing as debris on the nozzle jet interface. This prevented reliable jetting and aligned with difficulties reported in previous literature attempts at processing biomacromolecules. Digital light processing (DLP) was proposed as an alternative printing method that negated the requirement for formulations to pass through an orifice, which widens applicable materials selection. Samples were fabricated, allowing encapsulation of model proteins lysozyme (LYZ, 14 kDa), bovine serum albumin (BSA, 66 kDa) and alkaline phosphatase (ALP, 160 kDa). A range of PEGDA were selected from 575 to 10,000 Mn for formulation preparation in an effort to vary release from the printed polymer matrix. Key properties characterised included swelling ratio, swelling rate and theoretical matrix mesh size approximations, as determined by applying modified Peppas-Merrill equations. It was found that the respective values of each increased as Mn of the PEGDA selected increased. Limited release (17.6 ± 5.2% and 13.3 ± 1.2%) of LYZ and BSA was achieved from PEGDA Mn 10,000 20% (w/v) samples with 1 mg/mL model protein loading. A separate study using BSA considered the influence of increasing loading to 5 mg/mL, whereby release was then achieved using formulations of PEGDA Mn 4000, 8000 and 10,000 at 20% (w/v). Increased availability of free protein allowed increased release, but this was still limited with maximum elution being 25.8 ± 1.4% occurring over 3 days before plateau. DLP enabled sample production, but the formulations utilised and reliance on swelling alone was unsuitable for achieving long-term release. Finally, hydrolytically labile macromers were prepared through Michael addition modification of the previously used PEGDAs with an inexpensive thiol; dithiothreitol (DTT). Samples were fabricated by mixed-mode (MM) photopolymerisation, whereby increasing the thiol to acrylate ratio (SH:acrylate) increased areas of the polymer network that would hydrolytically degrade. Tuning the starting PEGDA Mn blend, macromer% (w/v) and SH:acrylate ratio allowed BSA release over 3 months+, fulfilling a main thesis aim. For example, samples using a PEGDA blend of Mn 575/4000 (1:1) 20% (w/v) and 0.25 SH:acrylate content with 5 mg/mL BSA loading led to release over 112 days with zero-order characteristics (R2 = 0.95) observed from day 2 of release onwards. Release of ALP was achieved with 56.6 ± 1.6% eluted over 21 days; an improvement on the PEGDA only formulations where no significant release was detected. Associated matrix mesh size approximations acted as a comparative tool between sample sets and provided reasoning for why differences in release occurred. Structural changes were confirmed using cryo-SEM imaging and mechanical properties assessed through compression testing in the as-printed and swelled (degraded) sample state. The results in this thesis suggest that kDa bioactives can be firstly encapsulated using simple scalable DLP printing processes and their elution controlled through changes in crosslinking mechanism, monomer selection, monomer%, SH:acrylate ratio, bioactive size and loading concentration. Throughout, emphasis has been placed on developing the understanding of the underlying principles associated with release of biomacromolecules from 3D printed structures. The information gathered could be applied to a further range of proteins in the development towards controlled delivery of protein therapeutics