Fuchs’ endothelial corneal dystrophy (FECD), first described by Austrian ophthalmologist Ernst Fuchs [
Usually, corneal endothelial abnormality in FECD may be observed before subjective symptoms develop [
In 2010, Shousha et al. [
Pentacam (OCULUS Optikgeräte GmbH, Wetzlar, Germany) is a rotating Scheimpflug camera, providing 360 degrees, three-dimensional, high-resolution images of the anterior segment structures from the cornea to the lens in a short acquisition time [
In this work, we aimed to evaluate the potential of Pentacam Scheimpflug densitometry as an objective, quantitative tool for approaching the morphological and optical alterations of cornea in FECD patients.
Institutional review board approval was obtained from the Chang Gung Medical Foundation, Taiwan. All data were collected from patients in the Chang Gung Memorial Hospital, Taiwan, during 2009–2012.
Study group consisted of the 27 patients (53 eyes) diagnosed with FECD by slit lamp biomicroscopy combined with corneal specular microscopy. Although the severity of FECD was usually graded clinically on the basis of the area and confluence of guttae, and the presence of edema [
All patients received the Pentacam examination under the same controlled ambient light conditions. Scheimpflug images were acquired automatically in 360 degrees fashion in the three-dimensional scan mode. Full-thickness corneal backscatter evaluation showed the measured maximum optical backscattering intensities on a densitogram on a relative scale from 0 to 100 (0 = no clouding, 100 = tissue completely opaque). A white marker line was positioned across the cornea and using the grid pattern superimposed on the image (Figure
Densitometry of Pentacam Scheimpflug images. (a) One spiking hump with central flattening and a smoothing second hump appear as a high-back chair in normal subjects. (b) Two spiking humps look like a hanging hammock in FECD patients.
Basic settings of six selected segments in Pentacam.
Statistical analysis was conducted using SPSS statistics 17.0 (SPSS Inc., Chicago, IL). To compare the difference between study and control groups, generalized estimating equation (GEE) was used.
Of total 106 Scheimpflug image studies, 53 were obtained from 27 FECD patients (aged 41–89 years, mean 65 years). The other 53 were from 27 normal subjects (aged 42–85 years, mean 60.78 years). Mean CCT measured with Pentacam were 572.42
Representative corneal densitograms of normal subject (Figure
Quantitative parameters in FECD and normal subjects.
FECD | Normal control | |||||
---|---|---|---|---|---|---|
CCT ( |
Mean AD (%) | DM/AD | CCT ( |
Mean AD (%) | DM/AD | |
Numbers | 23 | 23 | 23 | 23 | 23 | 23 |
Median | 576 | 18.10 | 1.13 | 551 | 16.10 | 0.91 |
Minimum | 511 | 13.65 | 0.74 | 505 | 13.13 | 0.79 |
Maximum | 1189 | 56.48 | 2.32 | 667 | 20.07 | 1.10 |
FECD: Fuchs’ endothelial corneal dystrophy; CCT: central corneal thickness; Mean AD: average area density; DM/AD: average ratio of Descemet’s membrane density versus area density.
Scatterplots of mean area density (mean AD) against corneal thickness (CCT) (a) and average ratio of DM density versus area density (DM/AD) (b) in FECD without edema, FECD with edema, and normal eyes.
Parameters as mean AD, DM/AD, and CCT were compared with GEE analysis for 99 eyes, including 53 normal eyes and 46 FECD with guttae alone eyes. It revealed no significant differences in CCT (Wald chi-square 0.270,
Parameters as mean AD, DM/AD, and CCT were compared for all 53 diseased eyes in the study group, which were further separated by clinical examination into guttae alone group (46 eyes in 24 patients) and edema group (7 eyes in 4 patients). GEE revealed no significant differences in CCT (Wald chi-square 0.261,
In this study, we used Pentacam densitogram to monitor the corneal backscattering effects in FECD patients. In qualitative analysis, the hanging-hammock pattern was observed on the densitograms in FECD patients, which was morphologically different from the high-back chair pattern in the normal subjects (Figure
Quantitative analyses revealed significant elevation of mean AD and DM/AD in FECD patients as compared to those in normal subjects. Mean AD is the average optical area density in the central 4 mm area of cornea, which increases in FECD, regardless of the disease condition. As it represents the increase of corneal backscatter in FECD, it may therefore imply that the optical quality of transparent corneal tissue was affected in most of FECD patients, regardless of the presence of corneal edema.
DM/AD is the proportion of DM backscattering effects on the entire corneal optical intensity changes, which also elevates in FECD. This indicates that except the backscattering effect caused by corneal edema, pathological change in DM alone also contributes to the backscattering increase in FECD. Therefore, in the early stage of FECD, when pathological changes occur only in DM, backscattering also augments. This can contribute to impairment of optical quality of the cornea in the early stage of FECD when corneal edema is not observed clinically. We think that the increased backscattering of DM in FECD might be related to the pathological changes of DM. In the early-onset form of FECD, the anterior band layer of DM thickened, and an additional internal layer of wide-spacing collagen containing type VIII collagen within posterior nonband layer is demonstrated. An additional posterior striated layer rich in type VIII, which is not present in normal cornea, also contributes to the thickening of DM in early-onset form of FECD [
Within the FECD group, we noticed the elevation of the densitogram curve with less depression between double humps in corneal edema patients. Hence, we tested the relation between the presence of corneal edema and our proposed parameters: CCT, mean AD, and DM/AD. We found that DM/AD failed to show significant difference between guttae alone group and edema group. This may be because of the limited case number and the variety of edematous degree we collected in edematous group.
As for the relation between corneal edema and CCT, the result showed no surprisingly irrelevance. This not only explains the individual anatomical variation in CCT but also demonstrates the advantage of using Pentacam rather than corneal pachymetry in monitoring the disease progression of FECD.
The only parameter that differs significantly between guttae alone and edema groups is the increased mean AD in the latter. In swollen cornea, the uneven distribution of fluid, the change in collagen fibrils arrangement, and the change in refractive index result in the increase of scattering in the cornea, [
In brief, in the early stage of FECD when the cornea is not edematous, diseased DM causes significant increase of corneal backscattering as compared to normal ones (different densitometric pattern, increased DM/AD, and increased mean AD). Along with the progression of FECD, corneal edema develops. The morphological pattern of corneal densitogram then changes, and the optical area backscattering augments (increased mean AD), which renders mean AD a better indicator than CCT considering the influence of optical quality in FECD corneas.
There are a few limitations in our work. Although Pentacam has already been equipped with advanced technology capable of producing high-resolution images, it is still not yet detailed enough in resolution compared to either spectral domain OCT or UHR-OCT. Nevertheless, Pentacam with its densitometry program quantifies the light backscattering effects, which approximates assessment of optical quality of cornea. Adding to that, Pentacam is more accessible in institutes or hospitals where corneal refractive surgeries are performed. This is a retrospective case-control pilot study. Visual acuity was not taken into analysis because its change involves multiple factors, including cataract condition. The Krachmer scale was not used to correspond our findings to it. Although the Krachmer scale is the most commonly used method to document the clinical severity in FECD, it is not always used by every clinician. While we were trying to test the validity of corneal backscatters measured by the Pentacam Scheimpflug imaging on assisting the FECD diagnosis, we could not have sufficient documentation with the Krachmer scale. Instead, we used the presence of corneal edema or not as a parameter to categorize FECD cases into “without edema, compatible with Krachmer’s grade 0–4” and “with edema, compatible with Krachmer’s grade 5.” Our work acts as a pilot study, trying to demonstrate the feasibility of the Pentacam Scheimpflug imaging as an objective and quantitative approach for FECD. Next, we will conduct a longitudinal study in order to assess the feasibility for corneal densitometry to monitor FECD patients. When we have a larger case number, we will try to identify specific cut-off values with receiver operating characteristic curve to determine sensitivity and specificity, to analyze the possible correlation between the 3D reconstructed Scheimpflug imaging of FECD corneas and Krachmer’s grading, and to consolidate the validation of these parameters for detection or grading of severity of FECD.
In conclusion, we demonstrated that corneal backscatters measured by the Pentacam Scheimpflug imaging system can be provided as an objective, quantitative index for assessing optical health of the cornea in FECD. It not only potentially provides adjunctive diagnostic/monitoring information with stronger quantitative evidence but also enables assessment of the optical quality in FECD corneas with different disease conditions. As Pentacam is readily accessible and is a quick noninvasive examination with much information provided, the densitogram program can assist ophthalmologists in objectively and quantitatively detecting the morphological/optical changes in diseased corneas. And corneal backscatters can therefore be provided as an effective index for corneal optical health in various corneal pathologies. This is a pilot study. Further studies are needed to consolidate the validation of these parameters.
The authors declare that there is no conflict of interest regarding the publication of this paper.