Purpose: Identifying pre/perioperative factors that predict corneal endothelial-cell loss (ECL) after phacoemulsification may reveal ways to reduce ECL. Our literature analysis showed that 37 studies have investigated one or several such factors but all have significant limitations. Therefore, the data of a large randomized controlled trial (PERCEPOLIS) were subjected to post-hoc multivariate analysis determining the ability of nine pre/perioperative variables to predict ECL.
Methods: PERCEPOLIS was conducted in 2015-2016 to compare two phacoemulsification techniques (subluxation and divide-and-conquer) in terms of 3-month ECL. Non-inferiority between the techniques was found. In the present study, post-hoc univariate and multivariate analyses were conducted to determine associations between ECL and age, sex, cataract density, preoperative endothelial-cell density, phacoemulsification technique, effective phaco time (EPT), and 2-hour central-corneal thickness. The data are presented in the context of a narrative review of the literature.
Results: Three-month data were available for 275 patients (94% of the randomized cohort; mean age, 74 years; 58% women). Mean LOCSIII cataract grade was 3.2. Mean EPT was 6 seconds. Mean ECL was 13%. Only an older age (beta = 0.2%, p = 0.049) and higher EPT (beta = 1.2%, p = 0.0002) predicted 3-month ECL. Cataract density was significant on univariate (p = 0.04) but not multivariate analysis. The other variables did not associate with ECL.
Conclusions: Older age may amplify ECL due to increased endothelial cell fragility. EPT may promote ECL via cataract density-dependent and -independent mechanisms that should be considered in future phacoemulsification research aiming to reduce ECL. Our literature analysis showed that the average ECL for relatively unselected consecutively-sampled cohorts is 12%.
Copyright: © 2024 Perone et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.