A Novel Endothelial Stress Index Combining Polymegethism And Pleomorphism: Correlation With Keratoconus Severity And Cross-Linking Response

Current keratoconus assessment relies on individual corneal endothelial parameters such as coefficient of variation (CV) and hexagonal cell percentage (HEX), each capturing only a single dimension of endothelial morphological stress. No composite metric exists that integrates polymegethism and pleomorphism into a unified biomarker. This study proposes and validates a novel Endothelial Stress Index (ESI), defined as CV/HEX×100, and evaluates its correlation with keratoconus severity staging, central corneal thickness, and maximum keratometry. Furthermore, we investigated whether baseline ESI can predict favorable topographic response following corneal cross-linking (CXL), potentially serving as a preoperative decision-support tool.

This study was conducted at the Department of Ophthalmology, Recep Tayyip Erdoğan University Training and Research Hospital, a tertiary referral center in Rize, Turkey.

In this prospective study, 60 keratoconus eyes and 60 age-matched healthy controls underwent specular microscopy (Tomey EM-4000). ESI was calculated as CV/HEX×100. Keratoconus eyes were staged per Amsler-Krumeich. Kmax and CCT were recorded. Patients underwent epithelium-off CXL with ESI reassessed at 1, 3, and 6 months. Univariate correlations used Pearson/Spearman tests. Multivariate logistic regression assessed ESI as an independent predictor of favorable CXL outcome (Kmax reduction ≥0.5 D at 6 months), adjusting for Kmax, CCT, age, and stage. ROC with DeLong comparison evaluated ESI versus individual parameters.

Mean baseline ESI was significantly higher in keratoconus vs controls (56.8±14.2 vs 46.1±8.7; p<0.001). ESI correlated with Kmax (r=0.52; p<0.001) and CCT (r=-0.38; p=0.003). ESI increased with Amsler-Krumeich stage (I: 48.3±9.1, II: 57.2±11.8, III: 68.4±15.6; p<0.001 for trend). On multivariate regression, ESI remained an independent predictor of favorable CXL response after adjusting for Kmax, CCT, age, and stage (OR=1.04, 95% CI: 1.01-1.08; p=0.018). ROC yielded AUC=0.72 for ESI; combining ESI with Kmax improved AUC to 0.81 (DeLong p=0.024). Neither CV (AUC=0.61) nor HEX alone (AUC=0.58) matched ESI. Post-CXL, ESI peaked at month 1 (65.2±16.8; p<0.001) and recovered by month 6 (58.1±14.5; p=0.412 vs baseline).

The proposed Endothelial Stress Index is a simple, reproducible composite metric that independently predicts CXL response beyond established parameters including Kmax and CCT on multivariate analysis. ESI captures combined morphological stress missed by individual endothelial measures, and its combination with Kmax yields strong predictive accuracy (AUC 0.81). The characteristic post-CXL ESI elevation with subsequent recovery provides a quantitative framework for monitoring endothelial safety. ESI may serve as a clinically practical preoperative and postoperative tool for keratoconus management, warranting multicenter validation.