Quantitative assessment of ophthalmic viscosurgical device retention using in vivo confocal microscopy.

Purpose: To develop and apply a new laboratory method for in vivo quantitative assessment of the retention of ophthalmic viscosurgical devices (OVDs) following phacoemulsification.

Setting: Department of Ophthalmology, University of Texas Southwestern Medical Center, Dallas, Texas, USA.

Methods: Studies of both eyes of New Zealand White rabbits were performed. Six OVDs were evaluated: Provisc and Healon (both sodium hyaluronate 1%), Healon5 (sodium hyaluronate 2.3%), Amvisc Plus (sodium hyaluronate 1.6%), Viscoat (sodium hyaluronate 3%-chondroitin sulfate 4%), and a new viscous-dispersive OVD, DisCoVisc (sodium hyaluronate 3%-chondroitin sulfate 4%). The OVD was injected to fill the anterior chamber and a phacoemulsification needle inserted with the tip positioned just anterior to the lens capsule. Simulated phacoemulsification was performed for 1 minute using flow rates of 20, 40, and 60 mL/min; a vacuum level of 300 mm Hg; and ultrasound power of 60% using a Legacy phacoemulsification unit. The needle was removed, and silicone oil (1000 centistokes) was injected into the anterior chamber. The distance between the corneal endothelium and the OVD-silicone oil interface was measured using in vivo confocal microscopy through-focusing (CMTF).

Results: Significant differences in residual thickness were found between the OVDs tested. Specifically, the residual thickness of both DisCoVisc (mean 324.5 microm +/- 163.7 [SD]) and Viscoat (251.4 +/- 100.9 microm) was significantly greater than that of Provisc (9.5 +/- 16.7 microm), Healon (3.8 +/- 11.3 microm), Healon5 (0.6 +/- 2.4 microm), and Amvisc Plus (65.6 +/- 134.0 microm) (P < .05, Dunn test). Ophthalmic viscosurgical device retention was greatest with DisCoVisc; however, there was no statistically significant difference between DisCoVisc and Viscoat in residual thickness. The flow rate did not have a significant effect on the residual thickness (Friedman 2-way analysis of variance by ranks).

Conclusions: Residual OVD thickness following simulated phacoemulsification could be quantitatively measured using in vivo CMTF. The results are consistent with human surgical experience in that the dispersive OVD (Viscoat) was better retained than the cohesive compounds. DisCoVisc, a new viscous-dispersive OVD, also showed retention compared with Viscoat under the experimental conditions.