Corneal thickness measurements with Scheimpflug and slit scanning imaging techniques in keratoconus

Reduced Retinal Blood Vessel Densities Measured by Optical Coherence Tomography Angiography in Keratoconus Patients Are Negatively Correlated with Keratoconus Severity

Keratoconus (KC) is the most common corneal ectasia. Optical coherence tomography angiography (OCT-A) is a relatively new non-invasive imaging technique that allows the visualization and quantification of retinal and choriocapillary blood vessels. The aim of this study is to assess retinal and choriocapillary vessel density (VD) differences between KC patients and healthy controls and to investigate correlations between VD and KC severity. Fifty-two eyes were included in this exploratory study: twenty-six eyes from 26 KC patients and twenty-six eyes from 26 age- and gender-matched healthy controls. All patients underwent Scheimpflug corneal topography with Pentacam, axis lengths measurement and optical coherence tomography angiography (OCT-A). The thinnest spot in corneal pachymetry, maximum K (Kmax) and KC severity indices from the Belin/Ambrósio enhanced ectasia display (BAD) were also assessed. There was a distinct reduction particularly in the retinal VD of the superficial capillary plexus (SCP). Correlation analyses showed strong and moderate negative correlations between the VD in the macular SCP and BAD KC scores and between the SCP VD and Kmax. There was no difference in retinal thickness between the KC and healthy controls. With this study, further evidence for altered VD measurements by OCT-A in KC patients is given. For the first time, we demonstrated negative correlations between BAD KC scores and retinal blood vessel alterations. A major limitation of the study is the relatively small sample size. Since an artefactual reduction of the quantitative OCT-A measurements due to irregular corneal topography in KC must be assumed, it remains to be investigated whether there are also actual changes in the retinal microcirculation in KC.

Agreement between anterior segment swept source-OCT and Scheimpflug imaging corneal aberration measurements in healthy eyes

PURPOSE

To assess agreement between corneal aberration measurements made through swept-source optical coherence tomography using a new anterior segment imaging device (Anterion) and a Scheimpflug imaging device (Pentacam HR) in healthy subjects.

METHODS

Cross-sectional study. In 50 eyes of 50 healthy subjects, 14 aberration parameters (7 across the anterior corneal surface and 7 across the total surface) were measured in 4 mm and 6 mm optic zones using each device: oblique trefoil (Z3_-3), vertical coma (Z3_-1), horizontal coma (Z3_1), horizontal trefoil (Z3_3), spherical aberration (Z4_0), root mean square (RMS) lower order aberrations (LOA) and RMS higher order aberrations (HOA). Data for the two devices were compared through intraclass correlation coefficients (ICC), paired t tests, limits of agreement (LoA) and Bland Altman plots.

RESULTS

Vertical coma was the only corneal aberration parameter that consistently showed excellent agreement (ICC > 0.8, mean difference -0.019, LoA -0.165 to 0.126). Good agreement (ICC = 0.75) between the devices was observed for RMS HOA, but this was slightly worse in the 6 mm optical zone (ICC = 0.667 for anterior RMS HOA). No over- or underestimation trend by one or other device was noted. Agreement was poor to moderate for the rest of the corneal parameters (ICC 0.2 to 0.7).

CONCLUSION

Despite good agreement overall for vertical coma and RMS HOA values, agreement for the remaining corneal aberration measurements was poor to moderate. As mean differences in our sample were overall small, in normal eyes these devices could be clinically judged as interchangeable.

Orbscan 3 Versus Pentacam HR: Evaluating the Possible Interchangeable Use of Various Parameters

PURPOSE

Many topography/tomography devices are available, and their interchangeable use is frequently questioned. This study aimed to compare the analogous indices of Orbscan 3 and Pentacam HR to detect whether the interchangeable use of the 2 devices is feasible.

METHODS

This retrospective study was conducted at Eye World Hospital, Giza, Egypt. Displays of Orbscan 3 (Bausch & Lomb; Technolas Perfect Vision GmbH) and Pentacam HR (Oculus Optikegrate) for 660 healthy right eyes were explored for various analogous indices. Bland-Altman plots with 95% limits of agreement (LoA) were used for comparisons.

RESULTS

The highest systematic error and lowest agreement existed in the front and back elevations from the best fit sphere (mean differences of 4.7 and 9.2 μm, respectively, and 95% LoA of -2.5 to 12.0 and -10.0 to 28.6, respectively). There was a relatively wide 95% LoA for each pachymetry at the corneal apex (-18 to 22 μm) and at the thinnest location (-22 to 22 μm), anterior chamber depth (-0.88 to 0.95 mm), flattest keratometry (-0.44 to 0.62 D), and steepest keratometry (-0.71 to 0.89 D). By contrast, the lowest systematic error and highest agreement existed for front and back best fit sphere radii of curvature (mean difference of 0.08 mm and 95% LoA of -0.08 to 0.24 and mean difference of 0.03 mm and 95% LoA of -0.17 to 0.23, respectively).

CONCLUSIONS

This study negates the accuracy of the use of devices interchangeably regarding most parameters. Therefore, examinations with the 2 devices should not be regarded comparable.

Sensitivity and Specificity of Belin Ambrosio Enhanced Ectasia Display in Early Diagnosis of Keratoconus

Purpose

Early diagnosis of keratoconus disease (KCN) is the first priority in the preoperative evaluations of refractive surgery (RS).The aim of this study was to investigate the correlation between findings of Belin Ambrosio enhanced ectasia display (BAD) software and conventional corneal imaging (Orbscan and topography) in the early diagnosis of KCN.

Methods

For conducting this cross-sectional study, a total of 1000 eyes were selected from 500 patients that underwent the myopic photorefractive keratectomy surgery and were compared in four study groups during the years 2017-2018. In group 1, all topography, Orbscan, and BAD criteria were normal (65.8%).In contrast, in Group 2, at least one of the topography or Orbscan criteria as well as at least one BAD criterion (12.6%) were abnormal. In Group 3, the eyes had normal Orbscan and topography criteria with at least one abnormal BAD criterion (18.5%). Also, in Group 4, the patients had at least one abnormal Orbscan or topography criterion, but all BAD criteria (3.1%) were normal. Thickness of the thinnest point (TP) of cornea was compared in Pentacam and topography. Data analysis was done by SPSS software (version 21).

Results

BAD criteria were normal in 78.5% of all eyes with normal topography and Orbscan criteria (specificity). BAD criteria were also abnormal in 80.2% of eyes (sensitivity). There was also no significant difference between TP in Orbscan and Pentacam.

Conclusions

BAD criteria had a relatively acceptable sensitivity and specificity, compared with conventional Orbscan and topography criteria. Thus, BAD criteria can be more effective in the early diagnosis of KCN.

Assessment of the Association between In Vivo Corneal Morphogeometrical Changes and Keratoconus Eyes with Severe Visual Limitation

Assessing changes suffered by the cornea as keratoconus progresses has proven to be vital for this disease diagnosis and treatment. This study determines the corneal biometric profile in eyes considered as affected by keratoconus (KC) showing severe visual limitation, by means of in vivo 3D modelling techniques. This observational case series study evaluated new objective indices in 50 healthy and 30 KC corneas, following a validated protocol created by our research group, which has been previously used for diagnosis and characterization of KC in asymptomatic (preclinical) and mild visually impaired eyes. Results show a statistically significant reduction of corneal volume and an increase of total corneal area in the severe KC group, being anterior and posterior corneal surfaces minimum thickness points the best correlated parameters, although with no discrimination between groups. Receiving operator curves were used to determine sensitivity and specificity of selected indices, being anterior and posterior apex deviations the ones which reached the highest area under the curve, both with very high sensitivity (96.7% and 90%, respectively) and specificity (94.0% and 99.9%, respectively). The results suggest that once severe visual loss appears, anterior corneal topography should be considered for a more accurate diagnosis of clinical KC, being anterior apex deviation the key metric discriminant. This study can be a useful tool for KC classification, helping doctors in diagnosing severe cases of the disease, and can help to characterize corneal changes that appear when severe KC is developed and how they relate with vision deterioration.

Corneal higher-order aberrations measurements: Comparison between Scheimpflug and dual Scheimpflug-Placido technology in keratoconic eyes

PURPOSE

To compare the difference and agreement of corneal higher-order aberrations (HOAs) in keratoconic eyes using Scheimpflug and dual Scheimpflug-Placido imaging systems.

SETTING

Emory University, Atlanta, Georgia, and the University of Southern California, Los Angeles, California, USA.

DESIGN

Retrospective between-devices reliability and agreement study.

METHODS

Patients diagnosed with keratoconus were evaluated sequentially by Scheimpflug and dual Scheimpflug-Placido devices. Differences, correlations, and agreement between values for total root mean square (RMS), trefoil, coma, and spherical aberration were analyzed, and Bland-Altman plots were generated.

RESULTS

Fifty eyes from 50 patients (31 men, 19 women) were evaluated. Trefoil at 30 degrees, spherical aberration, and total RMS were significantly different between groups (P < .05), whereas trefoil at 0 degrees and total coma values were not statistically different. There was a weak positive correlation between devices for trefoil at 0 degrees (r = 0.228), and a moderate positive correlation for trefoil at 30 degrees (r = 0.473), horizontal coma (r = 0.430), and for total corneal RMS (r = 0.637). Vertical coma (r = 0.816) and spherical aberration (r = 0.874) showed a strong positive correlation. The 95% limits of agreement (LoA) for absolute values were 1.963 μm for trefoil at 30 degrees, 2.449 μm for trefoil at 0 degrees, 3.530 μm for horizontal coma, 2.145 μm for vertical coma, 1.242 μm for spherical aberration, and 10.527 μm for RMS.

CONCLUSION

Significant differences were found between measurements of corneal HOAs generated by Scheimpflug and dual Scheimpflug-Placido devices in patients with keratoconus, with generally limited correlations and wide LoA. HOAs measurements from these devices should not be considered equivalent.

Comparing Change in Anterior Curvature After Corneal Cross-linking Using Scanning-slit and Scheimpflug Technology

PURPOSE

To evaluate the correlation between anterior axial curvature difference maps following corneal cross-linking (CXL) for progressive keratoconus obtained from Scheimpflug-based tomography and Placido-based topography.

DESIGN

Between-devices reliability analysis of randomized clinical trial data.

METHODS

Corneal imaging was collected at a single-center institution preoperatively and at 3, 6, and 12 months postoperatively using Scheimpflug-based tomography (Pentacam; Oculus Inc, Lynnwood, Washington, USA) and scanning-slit, Placido-based topography (Orbscan II; Bausch & Lomb, Rochester, New York, USA) in patients with progressive keratoconus receiving standard protocol CXL (3 mW/cm² for 30 minutes). Regularization index (RI), absolute maximum keratometry (K Max), and change in K Max (ΔK Max) were compared between the 2 devices at each time point.

RESULTS

Fifty-one eyes from 36 patients were evaluated at all time points. Values were significantly different at all time points (56.01 ± 5.3 diopters [D] Scheimpflug vs 55.04 ± 5.1 D scanning-slit preoperatively [P = .003]; 54.58 ± 5.3 D Scheimpflug vs 53.12 ± 4.9 D scanning-slit at 12 months [P < .0001]) but strongly correlated between devices (r = 0.90-0.93) at all time points. The devices were not significantly different at any time point for either ΔK Max or RI but were poorly correlated at all time points (r = 0.41-0.53 for ΔK Max, r = 0.29-0.48 for RI). At 12 months, 95% limits of agreement were 7.51 D for absolute K Max, 8.61 D for ΔK Max, and 19.86 D for RI.

CONCLUSIONS

Measurements using Scheimpflug and scanning-slit Placido-based technology are correlated but not interchangeable. Both devices appear reasonable for separately monitoring the cornea's response to CXL; however, caution should be used when comparing results obtained with one measuring technology to the other.