Scleral Thickness in Eyes with Pachychoroid Pigment Epitheliopathy Accompanied by Keratoconus

Purpose: To assess the prevalence of pachychoroid pigment epitheliopathy (PPE) in eyes with keratoconus (KC) and investigate its correlation with corneal, choroidal, and scleral indices with multimodal imaging. Design: An exploratory, cross-sectional, cohort study. Subjects: One hundred consecutive patients affected with KC. Main Outcome Measures Scleral stromal thickness, PPE prevalence, and their associations with corneal and choroidal parameters. Methods: Demographic data, corneal collagen cross-linking, anamnestic records, and clinical findings were collected. Imaging protocol included OCT (Spectralis HRA+OCT; Heidelberg Engineering), corneal topography (TMS-4N, Tomey), corneal pachymetry (RTVue-XR Avanti, Optovue), and axial length (AXL) measurement (OA-2000, Tomey). Anterior scleral stromal thickness was measured in the horizontal gaze positions 6 mm posteriorly to the scleral spur (Spectralis HRA+OCT; linear 20 degrees scan, 1024 A-scan per second). Odds ratios (ORs) and corresponding 95% confidence limits (95% CLs) were estimated through logistic regression analysis to evaluate the association between each study parameter and PPE. To accommodate for the potential clustering effect due to within-patient correlated eye data, a generalized estimating equation procedure was applied to regression analysis. Additionally, a decision tree machine learning model with K-fold cross-validation was employed to predict PPE. Results: Eighty-five Caucasian patients were eligible for analysis (mean age: 34.2 years, standard deviation: 8.7). The prevalence of PPE was 10.5% (95% CL: 4.9/19.1%; 9/85 patients; 11/170 eyes; 2 bilateral cases). Significant predictors for PPE according to logistic regression were choroidal thickness (OR: 4.51; 95% CL: 1.50/13.6 for 50 mu m increments; P = 0.007), age (OR: 4.61; 95% CL: 1.30/16.4 for 10-year increments; P = 0.018), and scleral stromal thickness (OR: 7.48: 95% CL: 1.69/33.1 for 25 mu m increments; P = 0.008). Sex, AXL, corneal curvature, and astigmatism parameters did not show significant discriminant ability (P > 0.05). Collagen cross-linking treatment was performed in a comparable proportion between the 2 groups (73.6% vs. 63.4% in PPE and non-PPE, respectively). Conclusions: Our study identifies increased scleral thickness as the key predictor of PPE in KC patients, followed by choroidal thickening and increased age. These findings provide new insights into the role of scleral biomechanics in KC eyes with PPE.