Abstract
Abstract: :
Purpose: To examine to what extent a vitrectomy affects retinal function and morphology, and to explore the effect of different vitreous cutting rates. Methods: Twelve rabbits were divided in two groups. A 2–minute core vitrectomy was performed at either 600 or 1200 cuts/min following a posterior vitreous detachment. Full–field and multifocal ERG (mfERG) were registered pre– and postoperatively at day 3, 28 and 58. One animal from each group was sacrificed after 10 respectively 28 days and the rest after 58 days. Staining with hematoxylin and eosin was performed as well as immunohistochemistry with antibodies against GFAP. In full–field ERG, the b–wave amplitudes for the rod, cone and combined response were measured, as well as the implicit time for light–adapted maximal cone response. The b–wave amplitudes and implicit times at the different time intervals were compared to the preoperative values using the Wilcoxon rank sum test. These comparisons were made within each of the 2 groups as well as with the groups averaged. In mfERG, the traces of the 5 lowest rows (corresponding to and somewhat inferior to the visual streak) were grouped, and the amplitude average was calculated. The averaged trace amplitudes from the different time intervals (pre– and postoperatively) and the 2 groups were compared using the Wilcoxon rank sum test. Results: Three days postoperatively, the b–wave amplitudes of the full–field ERG in all conditions were significantly reduced. Four weeks postoperatively, a significant reduction in the rod response was still evident. At 8 weeks, all amplitude responses were again normal. The implicit time remained unchanged at all time intervals. No correlation was found between vitreous cutter speed and ERG response. No change in mfERG derived amplitudes was found after vitrectomy.Hematoxylin and eosin stained slides revealed a normal retinal architecture in all eyes. GFAP labelling was detected in Müller cells of the entire retina in all eyes 3 days after surgery. At 29 days, the expression remained high in the eye in which the 600 cuts/min probe had been used, but was not detected in the 1200 cuts/min eye. At 58 days, a low level of Müller cell GFAP was detected in 3 of the 4 600 cuts/min. eyes, and in 1 of 4 1200 cuts/min. eyes. Conclusions: Vitrectomy transiently affects the retinal function, and rod photoreceptors are more affected than cones. Loss of function is well correlated to GFAP upregulation in retinal Müller cells, indicating a possible regulative effect of Müller cells on the retinal neurons. The use of high–speed vitrectomy reduces retinal trauma.
Keywords: vitreoretinal surgery • electroretinography: non-clinical • retina