Systematic assessment of microneedle injection into the mouse cornea

BACKGROUND: Corneal intrastromal injection is an important mode of gene-vector application to subepithelial layers. In a mouse model, this procedure is substantially complicated by the reduced corneal dimensions. Furthermore, it may be difficult to estimate the corneal area reached by the volume of a single injection. This study aimed to investigate intrastromal injections into the mouse cornea using different microneedles and to quantify the effect of injecting varying volumes. A reproducible injection technique is described. METHODS: Forty eyes of 20 129 Sv/J mice were tested. India ink was intrastromally injected using 30° beveled 33 G needles, tri-surface 25° beveled 35 G needles, or hand-pulled and 25° beveled glass needles. Each eye received a single injection of a volume of 1 or 2 μL. Corneoscleral buttons were fixed and flat mounted for computer-assisted quantification of the affected corneal area. Histological assessment was performed to investigate the intrastromal location of the injected dye. RESULTS: A mean corneal area of 5.0 ±1.4 mm(2) (mean ± SD) and 7.7 ±1.4 mm(2) was covered by intrastromal injections of 1 and 2 μL, respectively. The mean percentage of total corneal area reached ranged from 39% to 53% for 1 μL injections, and from 65% to 81% for 2 μL injections. Injections using the 33 G needles tended to provide the highest distribution area. Perforation rates were 8% for 30° beveled 33 G needles and 44% for tri-surface beveled 35 G needles. No perforation was observed with glass needle; however, intrastromal breakage of needle tips was noted in 25% of these cases. CONCLUSIONS: Intracorneal injection using a 30° beveled 33 G needle was safe and effective. The use of tri-surface beveled 35 G needles substantially increased the number of corneal perforations. Glass needles may break inside the corneal stroma. Injections of 1 μL and 2 μL resulted in an overall mean of 49% and 73% respectively of total corneal area involved.