Considerations in the sterile manufacture of polymeric microneedle arrays.

We describe, for the first time, considerations in the sterile manufacture of polymeric microneedle arrays. Microneedles (MN) made from dissolving polymeric matrices and loaded with the model drugs ovalbumin (OVA) and ibuprofen sodium and hydrogel-forming MN composed of "super-swelling" polymers and their corresponding lyophilised wafer drug reservoirs loaded with OVA and ibuprofen sodium were prepared aseptically or sterilised using commonly employed sterilisation techniques. Moist and dry heat sterilisation, understandably, damaged all devices, leaving aseptic production and gamma sterilisation as the only viable options. No measureable bioburden was detected in any of the prepared devices, and endotoxin levels were always below the US Food & Drug Administration limits (20 endotoxin units/device). Hydrogel-forming MN were unaffected by gamma irradiation (25 kGy) in terms of their physical properties or capabilities in delivering OVA and ibuprofen sodium across excised neonatal porcine skin in vitro. However, OVA content in dissolving MN (down from approximately 101.1 % recovery to approximately 58.3 % recovery) and lyophilised wafer-type drug reservoirs (down from approximately 99.7 % recovery to approximately 60.1 % recovery) was significantly reduced by gamma irradiation, while the skin permeation profile of ibuprofen sodium from gamma-irradiated dissolving MN was markedly different from their non-irradiated counterparts. It is clear that MN poses a very low risk to human health when used appropriately, as evidenced here by low endotoxin levels and absence of microbial contamination. However, if guarantees of absolute sterility of MN products are ultimately required by regulatory authorities, it will be necessary to investigate the effect of lower gamma doses on dissolving MN loaded with active pharmaceutical ingredients and lyophilised wafers loaded with biomolecules in order to avoid the expense and inconvenience of aseptic processing.