Corneal Backscatter Measures Edema Severity in Fuchs Dystrophy Cross Sectional Study

Optical measurements of corneal light backscatter increase progressively alongside the severity of corneal edema in patients diagnosed with Fuchs endothelial corneal dystrophy, providing an objective biomarker to help evaluate disease stages. In a cross-sectional clinical investigation evaluating ninety-nine eyes across sixty-seven individuals, researchers determined that total corneal backscatter in the central two-millimeter zone was significantly elevated among eyes presenting with higher stages of edema. Fuchs endothelial corneal dystrophy is widely recognized as the most common corneal dystrophy, characterized pathologically by progressive endothelial dysfunction that causes fluid accumulation and disrupts corneal transparency. Staging the condition accurately remains essential for monitoring disease progression and planning therapeutic interventions, yet conventional grading frameworks frequently rely on subjective slit-lamp observations that can show notable variability between different observers.

The study protocol received formal approval from the Institutional Review Board and Ethics Committee at Ruprecht-Karls University Heidelberg in Germany under identification S-621/2021, adhering strictly to the ethical tenets outlined in the Declaration of Helsinki with informed consent secured from all enrolled subjects. Eligible participants were examined between 2020 and 2025 by the same cornea consultant, ensuring standardized clinical evaluations across the study period. A total of one hundred and eight patients underwent initial screening, following which forty-one patients were excluded because they failed to meet inclusion criteria in both eyes. Among the remaining sixty-seven eligible patients, thirty-five individual eyes were excluded from the analysis due to coexisting ocular comorbidities in twenty eyes or insufficient tomographic scan quality in fifteen eyes, leaving a final cohort of ninety-nine eyes for comprehensive statistical assessment.

The study protocol established precise classification criteria to separate the ninety-nine eyes into three discrete cohorts based on their swelling status: thirty-one eyes presented without corneal edema, thirty-seven eyes exhibited subclinical corneal edema, and thirty-one eyes displayed clinically evident corneal edema. Baseline disease staging incorporated slit-lamp biomicroscopy according to the modified Krachmer scale, enrolling eyes displaying grade five confluent guttae exceeding five millimeters in diameter without clinically visible edema alongside eyes displaying grade six confluent guttae accompanied by clinically visible stromal swelling. Corrected distance visual acuity was recorded systematically in the morning under standardized photopic illumination calibrated at 320 candelas per square meter using an electronic five-letter per-line acuity chart positioned at a five-meter testing distance. For eyes lacking overt slit-lamp findings of fluid retention, subclinical tomographical edema was defined according to established tomographic criteria, requiring the positive identification of at least two of three distinct features: loss of regular parallel isopachs, spatial displacement of the thinnest corneal point away from the center, and focal depression of the posterior corneal surface.

All tomographic acquisitions were performed using a Pentacam AXL Scheimpflug imaging device, which generated standardized four-maps refractive analyses alongside corneal densitometry reconstructions. The Scheimpflug software automatically segmented the cornea into geometric regions across both horizontal and depth dimensions. Depth-resolved analyses partitioned the tissue into an anterior layer encompassing the superficial 120 micrometers, a posterior layer comprising the innermost 60 micrometers adjacent to the endothelium, and an intermediate central layer located between them. Transverse planar analyses were structured as concentric annular zones centered on the visual axis, comprising the central 0 to 2 millimeter zone, followed by rings spanning 2 to 6 millimeters, 6 to 10 millimeters, and 10 to 12 millimeters, as well as an integrated total diameter measurement spanning 0 to 12 millimeters. Central corneal thickness and thinnest corneal thickness were extracted concurrently from the refractive maps.

To evaluate the primary outcome regarding total corneal backscatter differences within the central two-millimeter zone across the three edema cohorts, the investigators conducted clustered Wilcoxon rank-sum tests applying the Datta-Satten method to correct for statistical interdependence arising from the inclusion of fellow eyes in bilateral participants. Accounting for three pairwise subgroup comparisons, the baseline significance threshold of 0.05 was adjusted to 0.017. Total central two-millimeter backscatter was significantly lower in the group without corneal edema, which exhibited a median value of 19.50 gray scale units with an interquartile range of 2.85, compared with the subclinical edema cohort, which demonstrated a median of 22.85 gray scale units with an interquartile range of 7.1, yielding a statistically significant difference at P equal to 0.006. The clinical edema cohort demonstrated a further elevated median of 26.35 gray scale units with an interquartile range of 11.03, which differed significantly from eyes lacking edema at P below 0.001 and from eyes with subclinical edema at P equal to 0.016.

Spatial and depth-resolved analyses revealed that the magnitude of backscatter elevation was highly localized rather than uniformly distributed across the entire corneal volume. The most pronounced numerical differences between the three edema categories occurred within the central two-millimeter zone and specifically within the superficial anterior layer measuring 120 micrometers in thickness. As measurement fields moved progressively outward toward the peripheral corneal annuli spanning from 2 to 6 millimeters, 6 to 10 millimeters, and 10 to 12 millimeters, median backscatter readings between the clinical subgroups converged and became increasingly indistinguishable. A corresponding attenuation of group differences emerged across tissue depth, where backscatter values recorded in the deeper posterior 60-micrometer layer showed far smaller variations between edema categories than those documented in superficial strata. Because patient age represented a potential confounding factor that might independently increase optical density, researchers performed mixed-effects modeling evaluating age across all annuli and layers, confirming that age lacked any statistically significant relationship with corneal backscatter in the examined cohort.

Secondary statistical analyses examined the clinical relevance of light backscatter by correlating tomographic measurements against corrected distance visual acuity using linear mixed-effects models clustered by patient identifier. Total central two-millimeter backscatter demonstrated a statistically significant moderate correlation with visual acuity, yielding a beta coefficient of 0.013, a 95 percent confidence interval from 0.007 to 0.018, a marginal R-squared of 0.208, and a Cohen f-squared effect size of 0.26 at P below 0.001. An even stronger relationship was documented for backscatter measured specifically within the anterior 120 micrometers of the central zone, exhibiting a beta coefficient of 0.009, a 95 percent confidence interval from 0.006 to 0.012, a marginal R-squared of 0.260, and a Cohen f-squared of 0.35 at P below 0.001. Notably, anterior backscatter correlated more strongly with visual acuity than established pachymetric indices, including central corneal thickness, which yielded a marginal R-squared of 0.198, and thinnest corneal thickness, which yielded a marginal R-squared of 0.096, while flat keratometry of the posterior surface maintained the highest individual correlation at a marginal R-squared of 0.304.

Diagnostic accuracy analyses using receiver operating characteristic curves characterized the capacity of Scheimpflug parameters to discriminate between adjacent disease categories based on optimal thresholds identified by the highest Youden index. In differentiating eyes without edema from eyes exhibiting subclinical edema, flat posterior keratometry achieved an area under the curve of 0.86 with an optimal cut-off of 6.77 millimeters, posterior asphericity reached an area of 0.81 at a threshold of minus 0.50, and central corneal thickness achieved an area of 0.76 at 571.5 micrometers. Central backscatter parameters yielded moderate discriminative accuracy, with the central layer reaching an area of 0.75 at 19.8 gray scale units, anterior backscatter reaching 0.73 at 31.2 gray scale units, and total central backscatter reaching 0.73 at 21.0 gray scale units. When distinguishing subclinical edema from overt clinical edema, anterior corneal backscatter emerged as the top-performing metric, producing an area under the curve of 0.76 at a threshold of 34.9 gray scale units with 78.1 percent sensitivity and 66.7 percent specificity, surpassing central corneal thickness at 0.75 and total central backscatter at 0.68.

To explore which structural indices independently predict categorical edema classification, investigators constructed an exploratory multivariate regression model utilizing generalized estimating equations with stepwise backward selection clustered by patient. Initial candidate variables included anterior, central, and posterior densitometry of the central two-millimeter zone, posterior corneal flat keratometry, posterior asphericity, thinnest corneal thickness, and central corneal thickness, with all continuous measures standardized to enable direct comparison across disparate units. The stepwise reduction eliminated central densitometry, posterior densitometry, posterior flat keratometry, posterior asphericity, and thinnest corneal thickness because their contributions were not statistically significant at P above 0.05. The final predictive model retained only two independent explanatory factors: anterior densitometry in the central two-millimeter zone, which yielded a standardized estimate of 0.201 at P below 0.01, and central corneal thickness, which yielded a standardized estimate of 0.611 at P below 0.001. Together, these two parameters yielded an overall categorical prediction accuracy of 54.5 percent across the three edema stages.

The study authors highlighted several methodological considerations and physiological hypotheses when interpreting the documented densitometric patterns. Additionally, subclinical stromal fibrosis secondary to chronic fluid accumulation may cause persistent backscatter changes that do not resolve immediately. The researchers emphasized that the retrospective, single-center design and the necessary exclusion of eyes with comorbid conditions or low-quality scans could introduce selection bias. Furthermore, the patient sample included both phakic individuals and pseudophakic individuals whose surgery took place at least three months prior, and the proportion of pseudophakic eyes increased in higher edema stages, presenting an uncontrolled factor that warrants validation in prospective cohorts incorporating visual quality metrics such as contrast sensitivity and straylight.

Subject of Research: Evaluation of corneal backscatter as an objective biomarker for edema severity in Fuchs endothelial corneal dystrophy

Article Title: Corneal backscatter as a biomarker for edema severity in Fuchs endothelial corneal dystrophy: a cross-sectional study

Article References: Friedrich, M., Son, H.-S., Turava, L.-G., Hammer, M., Blöck, L., Chychko, L., Yildirim, T. M., Auffarth, G. U., & Augustin, V. A. (2026). Corneal backscatter as a biomarker for edema severity in Fuchs endothelial corneal dystrophy: a cross-sectional study. Eye and Vision, 13(1), Article 17. https://doi.org/10.1186/s40662-026-00483-3

Image Credits: AI Generated

DOI: 10.1186/s40662-026-00483-3

Keywords: Corneal Backscatter, Fuchs Endothelial Corneal Dystrophy, Scheimpflug Tomography, Corneal Edema, Visual Acuity, Corneal Densitometry, Central Corneal Thickness, Subclinical Edema, Corneal, backscatter, biomarker, edema