Evaluating the Global Consensus on Keratoconus and Ectatic Diseases Agreements Reached on Subclinical Keratoconus





Purpose


To evaluate the consensus agreements reached in the Global Consensus on Keratoconus and Ectatic Diseases project for subclinical keratoconus, specifically that posterior corneal elevation abnormalities must be present to diagnose mild or subclinical keratoconus.


Design


Literature review.


Methods


Database review (PubMed) was performed on January 2, 2024, to identify studies evaluating the ability of posterior corneal surface metrics to identify subclinical keratoconus using the following search terms: posterior corneal elevation; keratoconus screening; corneal ectasia; subclinical keratoconus; keratoconus suspect; and asymmetric keratoconus. Articles were included for final analysis if they evaluated the ability of the posterior corneal surface to identify subclinical keratoconus compared to anterior corneal surface and/or thickness metrics and reported area under the receiver operating curve (AUROC) data for multiple variables to allow for metric comparison. Metrics evaluated in each manuscript were categorized as anterior surface, thickness, or posterior surface. The relative discriminative performance of anterior surface (A), thickness (T), posterior surface (P), and the multimetric D score (D) metrics were evaluated based on AUROC, sensitivity, and specificity in differentiating subclinical keratoconus from normal controls were evaluated.


Results


There were 29 articles identified that met the inclusion criteria and were evaluated. In intrastudy comparison, anterior surface metrics (37.9%) and thickness metrics (39.2%) performed best at differentiating subclinical keratoconus from normal corneas, while only 4 out of 29 studies (13.8%) reported posterior metrics outperforming all metrics. In the subgroup analysis including the multimetric D score ( n = 15), anterior surface metrics performed best (33.3%), followed by the D score (26.7%). In this D subgroup, no paper reported superior posterior metric performance.


Conclusions


In aggregate, posterior corneal surface metrics performed worse than anterior corneal and thickness metrics in differentiating subclinical keratoconic eyes from normal controls. These results demonstrate a lack of evidence to support the consensus claim that posterior elevation abnormalities must be present to diagnose subclinical keratoconus.


K eratoconus detection at the earliest possible stage has significant implications for the preservation of visual function in patients with progressive ectatic disease , and for the safest performance of corneal refractive surgery by screening out candidates who are at increased risk for developing postoperative corneal ectasia. , This is particularly important as the prevalence of visually significant myopia is rising worldwide and is expected to surpass 50% in most countries by 2050, while keratoconus prevalence has risen as diagnostic techniques have advanced.


In 2015, four supranational corneal societies (Asia Cornea Society, Cornea Society, EuCornea, and PanCornea) utilized a modified Delphi technique, with panels of individuals selected by assigned coordinators, to arrive at consensus opinions regarding keratoconus and other ectatic diseases, which culminated in the Global Consensus on Keratoconus and Ectatic Diseases report. Three panels were formed, including one specifically tasked with keratoconus definition/diagnosis. Consensus was achieved when two thirds of panel members reached the same opinion.


Consensus agreements reached included the following:



  • •

    Abnormal posterior elevation and abnormal corneal thickness distribution are mandatory to diagnose keratoconus.


  • •

    Posterior corneal elevation abnormalities must be present to diagnose mild or subclinical keratoconus.



The report did not provide any specific data to support these agreements, nor did it provide a list of the references used to reach the consensus. Given the controversy surrounding this topic and the need for greater understanding of its clinical implications, the purpose of this paper was to conduct an extensive literature review to evaluate the accuracy of the consensus statements regarding the identification of subclinical keratoconus (SKC), specifically the claim that “posterior corneal elevation abnormalities must be present to diagnose mild or subclinical keratoconus.”


METHODS


Literature review was performed to identify published studies that could serve to evaluate the relative efficacy of posterior corneal surface metrics in the identification of subclinical keratoconus. As this project included only literature review, no IRB approval was required. To generate the article list for evaluation, PubMed database review was performed on January 2, 2024, using the following search terms: posterior corneal elevation; keratoconus screening; corneal ectasia; subclinical keratoconus; keratoconus suspect; and asymmetric keratoconus. This resulted in 1,245 identified results. This list was screened by article title for potential relevance for subclinical keratoconus identification by one of the authors ( Figure 1 ). Articles were excluded from further review if their focus was any of the following topics: endothelial disease, cataract surgery, corneal transplantation, or other topics not related to keratoconus. Articles with a primary focus on corneal cross-linking that did not evaluate keratoconus detection were also excluded at this juncture. The remaining article list was then cross-referenced with two recent review articles by Moshirfar and colleagues regarding tomography for refractive surgery screening , as well as the Ophthalmic Technology Assessment paper on advanced corneal imaging in keratoconus by the American Academy of Ophthalmology. These additional identified publications were again screened by title. After this step, 397 papers were further evaluated by abstract review to determine if they pertained to keratoconus imaging or refractive surgery screening. After abstract review, 212 articles were accessed and reviewed in detail. Articles were included for final analysis if they met all of the following criteria:



  • 1.

    Study included a group that could be considered “subclinical keratoconus” regardless of the specific terminology used (ie, forme fruste keratoconus, keratoconus suspect, asymmetric keratoconus, subclinical keratoconus, or other).


  • 2.

    Evaluated the ability of the posterior corneal surface to identify subclinical keratoconus compared to anterior corneal surface and/or thickness metrics.


  • 3.

    Reported area under the receiver operating curve (AUROC) data for multiple variables to allow for intrastudy metric comparison.




FIGURE 1


Flow diagram showing the studies selection process.


This final search resulted in 29 articles that met these criteria and were analyzed.


SUBCLINICAL KERATOCONUS CRITERIA and CLASSIFICATIONS ACROSS STUDIES


The term subclinical is typically defined as a disease state not yet sufficiently advanced to present definite or readily observable symptoms. In reviewing the literature, however, there were multiple disparate methodologies used to classify eyes as subclinical rather than clinical keratoconus. We did not create any unique classifications in this study but merely identified the criteria reported in each manuscript. In general, these classification systems could be categorized as the fellow eye of a patient with manifest keratoconus in one eye and: 1) clear distinct abnormalities on the anterior corneal surface that are well described in the Methods section that served as inclusion for SKC eyes and exclusion for control eyes (eg, inferior–superior asymmetry and/or a bow-tie pattern with skewed radial axes detected on tangential Placido disk–based topography, or KISA index between 60% and 100%, ), or 2) indistinct anterior curvature findings described in the Methods section that would thus not necessarily differentiate normal controls from subclinical eyes (eg, no topographic or clinical signs of keratoconus in one eye).


METRIC SUBGROUP CLASSIFICATIONS


A variety of individual metrics were assessed across studies; however, few studies reported efficacy of all potential metrics available from the imaging device used, and numerous studies focused only on select metrics for evaluation. Due to this heterogeneity of individual metrics reported from each study that limited individual metrics comparisons across studies, we categorized metrics evaluated in each manuscript into three subgroups: anterior surface metrics (A), thickness metrics (T), and posterior surface metrics (P). One parameter, the final “D” score, from the Pentacam device (Oculus), was also reported in numerous studies. As this metric is a multimetric variable utilizing anterior surface, thickness, and posterior surface data, it was evaluated as an independent metric in our analysis. We also identified articles that utilized only the final D score in their evaluations. These articles were not included in comparative data analysis. But were used to evaluate the efficacy of the D metric over time.


When determining which metric performed best in each paper, we used the reported AUROC values followed by sensitivity values. AUROC values within 0.03 were considered equivalent unless there was > 5% difference in sensitivity, in which case the more sensitive metric with equivalent AUROC was considered the best at identifying subclinical keratoconus and distinguishing between normal and subclinical groups. We did not perform a de novo statistical analysis on this data set but rather relied on the AUROC, sensitivity, and specificity values reported in each study.


Given the heterogeneity of metrics evaluated and definitions used to define the subclinical cohorts, it was not feasible to combine data from different studies. Rather, we evaluated the relative efficacy of the metrics reported within each study, aggregated by anterior surface, thickness, or posterior surface, and reported the best metric category (anterior, posterior, or thickness) for each article separately. This methodology allowed us to compare the best performing A, T, and P metric for each study regardless of the individual metrics utilized in that study and aggregate results based on subgroup efficacy.


RESULTS


We identified 29 articles that met our criteria and were analyzed. There were fifteen papers within this cohort that also reported AUROC data for the final D score value from the Pentacam device. , , , , , We also identified ten articles that utilized only the final D score in their evaluations. Among the 29 articles analyzed, only six studies (20.6%) used corrected distance visual acuity (CDVA) of 20/20 or better as an inclusion criteria for subclinical keratoconus. , , ,


Table 1 shows the various metrics that performed best at identifying subclinical keratoconus in one or more studies organized by A, T, or P subgroup categories. Multiple studies identified the same metric/s as being most effective in their studies. The final list of articles included for analysis is shown in Table 2 . The best performing anterior surface, thickness, and posterior surface metric within each article, with their corresponding AUROC, sensitivity, and specificity values, are listed as well as the same values for the overall D score for papers that included this metric. Studies that used only the final D score in their analysis are listed in Table 3 . Among these studies, only two (20%) used CDVA of 20/20 or better as an inclusion criteria for subclinical keratoconus. ,



TABLE 1

Best Subcategory Metrics Reported Across Literature Cohort for Identifying Subclinical Keratoconus
































Anterior Thickness Posterior
Curvature Absolute Values Curvature
Astig
Ks
Km
Kf a
Kmax (T)		 CTmin
CCT
CCT diff		 Ks (p)
Rc (p)
Asymmetry indexes Pachymetry indexes Asymmetry indexes
ISV
I-S
IVA
IHD
SRI a
KI		 PPImax
ARTmax
PPIavg
ARTavg
ARTmin
Dt		 AAI-P b
Elevation Location Elevation
AE BFS (t)
AE BFS (m)
AE BFTE (t)
AE BFS loc Z a 		Thin loc Y
Thin loc X		 BED
PE BFS (t)
PE BFS (m)
Radius post. BFS a
PE BFS loc Z a
Db

a Metrics obtained specifically from Galilei dual Scheimpflug system (Ziemer Ophthalmic System AG, Port, Switzerland).


b Metric obtained specifically from Sirius device (C.S.O, Italy). Anterior: AE BFS = anterior elevation best fit sphere; AE BFS loc Z = AE BFS in the z direction; AE BFS (t) = AE BFS at the thinnest location; AE BFS (m) = AE BFS at maximum elevation; AE BFTE (t) = AE best fit toric ellipsoid at the thinnest location; Astig = central astigmatism from the anterior corneal surface; IHD = index of height decentration; ISV = index of surface variance; I-S: inferior-superior asymmetry; IVA: index of vertical asymmetry; Kf = flat keratometry; KI = keratoconus index; Km = mean keratometry; Kmax (T) = maximum tangential keratometry; Ks = steep keratometry; SRI = surface regularity index. Posterior: AAI-p = posterior asymmetry and asphericity index; BED = back elevation difference; BFS = best fit sphere; BFTE = best fit toric ellipsoid; Db = deviation of normality of the back elevation; Ks (p): posterior curvature-steep keratometry; PE BFS (m) = maximum posterior elevation on the best-fit sphere; PE BFS (t) = PE BFS at the thinnest location; PE BFS loc Z = PE BFS in the z direction; Rad Post BFS = BFS radius of curvature; Rc (p) = posterior radius of curvature. Thickness: ART = Ambrósio relational thickness; ARTavg = average ART; ARTmax = ART maximum; ARTmin = minimum ART; CCT = corneal thickness at the apex of the cornea; CCTdiff = difference from the corneal thickness at the apex and the thinnest point; CTmin = thinnest corneal thickness; Dt = deviation of normality of corneal thinnest point at BAD; PPI = pachymetric progression index; PPIavg = average PPI; PPImax = maximum PPI; Thin loc X = location of the thinnest point in the X axis (horizontal); Thin loc Y = location of the thinnest point in the Y axis (vertical).



TABLE 2

List of Published Studies that Evaluated the Efficacy of the Posterior Corneal Surface in Identifying Subclinical Keratoconus


































































































































































































































































































































































































































































































































Yr/Ref Anterior Best AUC Sens Spec Thickness Best AUC Sens Spec Posterior Best AUC Sens Spec D AUC Sens Spec
2011 ISV 0.80 86 67 PPI max 0.84 84 78 PE BFS 0.79 82 67
2012 N/S N/S N/S N/S CTmin 0.95 91 89 PE BFTE 0.91 83 90
2013 Ks 0.79 Thin loc Y 0.82 PE BFS 0.92
2013 AE BFS 0.90 84 86 PE BFS 0.92 80 94
2013 Kmax (T) 0.86 84 70 PPImax 0.81 78 65 BED 0.76 74 65
2014 I-S 0.80 86 74 N/S N/S N/S N/S BED 0.74 57 88 N/S N/S N/S
2014 Ks 0.68 96 39 CTmin 0.79 70 80 Ks 0.64 35 75
2015 AE BFS 0.77 71 82 CTmin 0.85 78 81 PE BFS 0.88 71 94
2015 I-S 0.61 ARTmin 0.74 68 73 BED 0.76 51 85 0.83 60 90
2015 AE BFS 0.68 55 68 ARTmax 0.73 69 69 PE BFS 0.70 64 63 0.71 66 66
2016 IVA 0.86 82 73 N/S N/S N/S N/S N/S N/S N/S N/S 0.86 81 73
2016 IHD 0.88 80 75 PPI 0.79 67 69 PE BFS 0.75 73 71
2017 Km 0.82 64 80 ARTmax 0.88 82 70 BED 0.60 53 80 0.67 71 60
2017 Astig 0.92 85 93 CCT 0.80 72 78 Ks 0.87 88 90
2017 AE BFS 0.81 72 73 CTmin 0.70 92 47 PE BFS 0.86 93 67
2018 IVA 0.84 92 53 PPImax 0.84 97 58 PE BFS 0.89 96 76 0.90 96 74
2018 AE BFS 0.94 96 70 CCTdiff 0.85 89 59 PE BFS 0.90 93 64 0.88 93 71
2018 ISV 0.75 53 87 CTmin 0.74 53 87 N/S N/S N/S N/S 0.75 80 67
2018 IHD 0.78 53 86 ARTmax 0.89 78 84 PE BFS 0.80 78 70 0.89 55 95
2018 IHD 0.79 CTmin 0.64 PE BFS 0.55 0.62
2019 ISV 0.88 80 80 CCTdiff 0.79 70 70 N/S N/S N/S N/S
2019 AE BFS 0.72 52 85 Thin loc y 0.78 71 75 PE BFS 0.76 61 87
2020 ISV 0.79 77 73 ARTavg 0.91 81 91 BED 0.76 73 80 0.93 84 94
2020 I-S 0.73 81 52 Thin loc y 0.76 65 78 AAI 0.83 92 58
2020 IVA 0.88 85 85 ARTmax 0.87 92 78 BED 0.74 69 67 0.95 100 90
2020 AE BFS 0.76 77 65 PPImax 0.81 82 69 PE BFS 0.78 75 64 0.86 85 67
2021 KI 0.85 72 84 Dt 0.74 82 53 Db 0.67 72 63 0.80 72 82
2021 AE BFS 0.94 87 93 PPImax 0.75 74 71 PE BFS 0.89 87 79 0.82 71 77
2022 AE BFTE 0.93 81 89 N/S N/S N/S N/S PE BFTE 0.84 60 91

Only gold members can continue reading. Log In or Register to continue

Related posts:

  1. Epidemiologic Study of Pediatric Uveitis and its Ophthalmic Complications Using the Korean National Health Insurance Claim Database
  2. Phase 2 Study of the Anti-High Temperature Requirement A1 (HtrA1) Fab Galegenimab (FHTR2163) in Geographic Atrophy Secondary to Age-Related Macular Degeneration
  3. Drugs Associated With Floppy Iris Syndrome: A Real-World Population-Based Study
  4. Neurofibromatosis Type 1 (NF1)–Related Ocular Signs: New Insights on Their Prevalence, Incidence, and Genotype-Phenotype Correlation in NF1 Children

Stay updated, free articles. Join our Telegram channel
Tags:
Jul 26, 2025 | Posted by in OPHTHALMOLOGY | Comments Off on Evaluating the Global Consensus on Keratoconus and Ectatic Diseases Agreements Reached on Subclinical Keratoconus

Full access? Get Clinical Tree
Get Clinical Tree app for offline access