3D-Printed membrane as an alternative to amniotic membrane for ocular surface/conjunctival defect reconstruction: An in vitro & in vivo study.

BACKGROUND: The aim of this study was to evaluate the surgical handling and clinical applicability of a specific 3D-printed membrane design fabricated using a gelatin, elastin and sodium hyaluronate blend for conjunctival reconstruction and compare it with amniotic membrane (AM), which is normally used in such surgeries.

METHODS: 3D printing technique was employed to fabricate the membrane based on gradient design. Prior to printing, rheometry was employed to optimize the ink composition. The printed membranes were then fully characterized in terms of physical and mechanical properties. In vitro viability, proliferation and adhesion of human limbal epithelial cells were assessed using MTT assay and scanning electron microscopy (SEM), respectively. Prior to in vivo experiment, surgical handling of each membrane was evaluated by three surgeons. In vivo evaluation was conducted through implanting the gelatin-based membranes and AM on induced conjunctival defects in rabbits (n = 8). Clinical observations, including epithelialization, inflammation severity, scar tissue formation and presence of granulation tissue, were recorded from day 1 through day 28. Histological examination was performed on all enucleated eyes on day 28. In addition to H&E staining, specific stains including Periodic Acid Schiff staining, Masson's Trichrome staining and immuno-histochemical staining for α-SMA were further used to assess goblet cell proliferation, healed sub-epithelial stroma and scar tissue formation and the presence of myofibroblasts, respectively.

RESULTS: Among all the examined compositions, a blend of 8% w/v gelatin, 2% w/v elastin and 0.5% w/v sodium hyaluronate was found to be appropriate for printing. The printed membranes had favorable optical characteristics (colorless and transparent), and the surgical handling was significantly easier compared to AM. Epithelial cells cultivated on the membranes indicated suitable viability and proliferation, and SEM images presented appropriate cell adhesion on the surface of the membranes. Clinical observations suggested similar epithelialization time (approximately 3 weeks) for both the membrane and AM grafted eyes but significantly lower levels of clinical inflammation in the membrane group from day 1 through day 28 (p = 0.01), which is a key advantage of using the printed membranes over the AM. Histological examination showed similar qualities in the healed epithelium in terms of cell morphology and cell layers. However, twice the density of goblet cells per 100 cells was observed in the gelatin-based membrane grafted group. Remnant of the degraded implant was seen in only 3 of the membranes, but in 7 of the AM grafted eyes. Inflammation and granulomatous reaction was significantly higher in sections containing the AM compared to membrane (p < 0.01 and p = 0.01, respectively). α-SMA staining was more evident, but not significantly different from the gelatin-based membrane, for the AM group (p = 0.25).

CONCLUSION: The designed gelatin-based membrane offers the necessary physical and mechanical characteristics needed for successful ocular surface/conjunctival defect construction and may be considered a promising alternative to AM due to a more predictable degradation pattern, higher goblet cell density on the healed epithelium, less inflammation and reduced scar tissue formation.