Short-term refractive and corneal astigmatism after canal-based microinvasive glaucoma surgery.

PURPOSE: To investigate changes in refractive and corneal astigmatism after Hydrus Microstent (Alcon Inc., Geneva, Switzerland) implantation during microinvasive glaucoma surgery. SETTING: One German medical center. DESIGN: Retrospective, non-randomized, observational cohort study. METHODS: Subjective refraction was obtained mandatorily and keratometry (Pentacam, OCULUS Optikgeräte GmbH, Wetzlar, Germany) optionally preoperatively and after 3 months (D90). Refractive astigmatism was analyzed using three-dimensional (3D) power vectors, including the spherical equivalent, M, the Jackson cross-cylinder projections to the 0° and 90° meridian (J0) and to the 45° and 135° meridian (J45), and the blurring power vector B. Corneal anterior and posterior surface astigmatism was investigated using the vectorial parameters J0, J45, and B. A 3D power vector time-course analysis of refractive astigmatism was performed using multivariate pairwise testing, and linear mixed-effects models, accounting for patient ID as a random effect, were used for the remaining analyses. RESULTS: A total of 90 ocular hypertension/glaucoma eyes were included. Sixty eyes were implanted with a Hydrus Microstent and phacoemulsified, and 30 eyes underwent implantation as a stand-alone procedure. Univariate analysis of the refractive cylinder showed no variation over time in any cohort (Phacoemulsification: CD0 = -1.3 ± 0.9 D vs. CD90 = -1.2 ± 1.2 D, P = 0.6; Stand-alone: CD0 = -1.6 ± 1.3 D vs. CD90 = -1.6 ± 1.2 D, P = 0.9), and no relevant time-by-group interaction was observed. Multivariate astigmatism analysis (M, J0, J45) indicated a significant change in the phacoemulsification cohort (P = 0.0003), in contrast to the stand-alone procedure cohort (P = 0.5). In the phacoemulsification cohort, B significantly declined from 2.9 ± 2.8 D to 1.2 ± 1.0 D (P < 0.0001) but remained stable in the stand-alone group (2.7 ± 3.5 D to 2.8 ± 3.4 D, P = 0.07). Corneal astigmatism analysis, based on 55 eyes from 37 patients, showed that the magnitudes of the J0 and J45 vectors remained stable over time for both the anterior (D0: J0 = -0.12 ± 0.07 D, D90: J0 = -0.21 ± 0.07 D, P = 0.12; D0: J45 = -0.05 ± 0.05 D, D90: J45 = -0.01 ± 0.05 D, P = 0.5) and posterior surfaces (D0: J0 = 0.12 ± 0.01 D, D90: J0 = 0.13 ± 0.01 D, P = 0.3; D0: J45 = -0.01 ± 0.01 D, D90: J45 = -0.01 ± 0.01 D; P = 0.9), without significant effects of time, group, or their interaction. CONCLUSIONS: The Hydrus Microstent is neutral in refractive astigmatism when used as a stand-alone procedure. In contrast, combined implantation with phacoemulsification can induce changes in refractive astigmatism. Keratometry analysis confirmed that both anterior and posterior corneal astigmatism remain stable, indicating that refractive changes are likely due to internal factors associated with lens exchange rather than the stent itself. Therefore, when implanted alone, the Hydrus Microstent is unlikely to compromise the results of prior refractive treatments.