Introduction
Keratoconus (KC) is a bilateral, asymmetric, inflammatory progressive ectatic corneal disorder characterized by steepening, thinning, and ectasia of the cornea resulting in irregular astigmatism and decreased visual acuity.[ 1 ] The prevalence of KC in India is as high as 2.3% compared to Western countries 0.43%.[ 2 , 3 ] Various visual functions including visual acuity and contrast sensitivity are affected in KC patients and thereby affecting their vision-related quality of life.[ 4 ]
The detection of subclinical KC (SBKC) is challenging as any single diagnostic tool or single parameter alone is not sufficient to detect early changes effectively.[ 5 ] Corneal topography and tomography are considered the gold standard in diagnosis of ectatic disorders and refractive surgery screening. Although, topographers are good at diagnosing KC, the sensitivity and specificity of most of its parameters in diagnosing the SBKC or form fruste KC (FFKC) is poor.[ 5 ] Recently, corneal biomechanical assessment has gained a lot of attention due to its ability in separating abnormal corneas such as KC and post-LASIK ectasia from normal corneas.[ 6 ] Biomechanical assessment with an ocular response analyzer (ORA) is poor in differentiating keratoconic suspect corneas from normal corneas.[ 7 ] The Corvis ST assesses the dynamic corneal response which provides better information in detecting abnormal corneas as compared to ORA.[ 8 ] The instrument has shown good repeatability[ 9 , 10 ] and good diagnostic ability in detecting ectatic eyes.[ 6 , 11 ] Metrics such as Corvis biomechanical index (CBI) with a cutoff of 0.5 had shown a specificity and sensitivity of 98.4% and 100%, respectively.[ 6 ] Furthermore, tomographic biomechanical index (TBI) with a cutoff of 0.29 had a specificity and sensitivity of 96% and 90.4%, respectively.[ 11 ] Moreover, the AUROC of TBI for detecting ectasia (0.996) is significantly higher than CBI (0.936) and Belin Ambrosio enhanced ectasia display total deviation value (BAD- D) (0.956).[ 11 ]
Biomechanical changes precede structural changes.[ 12 ] In KC, collagen structure, pattern, and arrangement are affected resulting in loss of structural integrity that triggers the weakness of biomechanical properties leading to focal weakening.[ 13 ] Integration of two diagnostic modalities improves the specificity and sensitivity of detecting abnormal corneas.[ 11 ] A significant reduction of epithelial thickness in KC has been reported previously.[ 14 ] Recently, the role of epithelial mapping as an adjunct tool in detecting SBKC has been described.[ 15 ] However, there is limited literature about utility of epithelial mapping in the diagnosis of spectrum of KC including SBKC and FFKC. Here, we evaluated the utility of epithelial mapping, CBI, and TBI in diagnosing the spectrum of KC.
Methods
This was a retrospective study where a spectrum of KC subjects who had undergone corneal epithelial thickness measurement using anterior segment optical coherence tomography (OCT) and corneal biomechanical parameter measurements using Corvis ST between July 2020 and July 2021 were enrolled. The study was approved by the Institutional Ethics Committee and adheres to tenets of the Declaration of Helsinki. KC subjects with an age between 11 and 50 years were enrolled. Subjects with ocular diseases, history of previous corneal surgery, corneal scars or hydrops, ocular trauma, ocular surface disorder, contact lens wear during the past 3 weeks, systemic disease, missing data, and poor scan quality were excluded from the current analysis.
Subjects
A total of 65 spectrums of KC eyes were included in the analysis. KC was classified using Belin ABCD classification/staging system available in Pentacam.[ 16 ] The new ABCD classification consists of 4 parameters each having five stages (0–4, where 1 indicates mild and 4 indicates severe). It uses anterior (parameter “A”) and posterior (parameter “B”) radius of curvature in the 3.0-mm zone centered on the thinnest location of the cornea, thinnest pachymetry in μm (parameter “C”), and manually entered distance best-corrected visual acuity (parameter “D”). A patient with abnormal anterior and posterior radius of curvature with corneal thinning (A1 and above and/or B1 and above and C1 and above) was considered clinical KC whereas a subject with normal anterior radius of curvature (A0) and abnormal posterior curvature (B1 and above) along with corresponding corneal thinning was considered SBKC. FF was considered topographically normal eyes of unilateral KC.
Corneal epithelial thickness measurement
Epithelial thickness measurements were made using Fourier-domain OCT Avanti (Optovue, Inc., Fremont, CA, USA) with corneal adaptor module (CAM). CAM captures the images at wavelength of 830 nm and transverse and axial resolution of 15 μ and 5 μ, respectively. Further, Pachymetry + Cpwr scan pattern was used which captures 8 radial scans, each with 1024 A-scans, over 6-mm diameter. Epithelial thickness map comprises 17 zones: (1) a central 2-mm zone, (2) 8 zones within an annulus between 2 and 5 mm, and (3) 8 zones within an annulus between 5 and 6 mm. Epithelial thickness at the thinnest point and standard deviation (SD) of epithelial thickness were taken into consideration.
Corneal tomography and biomechanic measurement
Corneal tomography and biomechanical measurements were made using Corvis® ST integrated with Pentacam HR (Oculus Optikgerate, Wetzlar, Germany). Corneal tomography readings were obtained using Pentacam HR which uses Scheimpflug technology whereas biomechanical parameters were measured using the recently introduced Corvis® ST which records corneal deformation to a defined air pulse using a high-speed Scheimpflug camera. The camera takes over 4300 images per second and 576 points per image over central 8 mm of cornea. It measures corneal deformation parameters at different phases (at applanation 1, applanation 2, and highest concavity) and composite corneal biomechanical parameters such as CBI and TBI.
Statistical analysis
Data were entered in Data were entered in MS Excel (Microsoft Corporation) and analyzed using Minitab Software (Minitab LLC, State University, PA, USA). Means and SD were calculated for continuous variables and proportions for the categorical variables. To understand the utility of various corneal parameters in diagnosing spectrum of keratoconus, cutoff values for epithelial thickness at the thinnest location and its standard deviation (SD) were considered 45 and 3 microns, respectively, CBI of 0.5 and TBI of 0.29 was considered. Descriptive statistics were described for each variable.
Results
Sixty-five eyes of 34 patients with a mean ± SD age of 30.73 ± 5.71 were included, out of which 45 eyes were KC, 10 eyes were SBKC, and 10 eyes were forme fruste KC. Table 1 describes the mean ± SD intraocular pressure, thinnest pachymetry, and keratometry.
Table 1: Mean±standard deviation intraocular pressure, thinnest pachymetry, and keratometry
In our keratoconic sample, epithelial mapping alone helped diagnose the 77.77% of cases, however, when combined with CBI, it helped diagnose (17.73%) additional cases (95.5%) and including TBI, it was useful in diagnosing 22.23% of cases additionally totaling up to 100.00%. Thus, biomechanical parameters helped detect the 22.23% of additional cases of KC as compared to epithelial mapping as a sole parameter. Similarly, in SBKC group, 40% of cases were detected by epithelial mapping alone, and when combined with CBI, it helped diagnose 30% of additional cases (70%) and TBI further helped diagnose 20% of additional cases (90%). We did not see any role of epithelial mapping in detecting FFKC cases whereas CBI helped diagnose 60% of cases and TBI helped detect 30% of cases additionally totaling up to 90%. Table 2 describes the utility of various corneal parameters for diagnosing the KC.
Table 2: The utility of various corneal parameters for diagnosing the spectrum of keratoconus
Discussion
Corneal topography is a primary tool to detect the KC; however, early detection of SB or FFKC is challenging. The prevalence of subclinical KC has been reported to be 6.8% in males and 2.8% in females[ 17 ] which is higher yet underreported as compared to KC. This could be attributed to the lack of single modality or tools to diagnose the SB or FFKC. Recent advances and newer modalities have enabled in better detection of spectrum of KC. Here, we discuss the utility of epithelial mapping, CBI, and TBI in detail.
Epithelial mapping is considered an important single tool in detecting KC; however, there is limited literature about its ability to detect SB and FFKC. Various instruments such as very high-frequency digital ultrasound (VHFDU), confocal microscopy, MS-39 (combination of topographer and tomographer), and SD-OCT are available for measuring epithelial thickness.[ 15 , 18–20 ] In the present study, we used SD-OCT to map epithelial thickness which is noninvasive, and thus, it reduces the risk of corneal compression or infection unlike seen with VHFDU and confocal microscopy. Corneal biomechanic measurements in a clinical setting are challenging. ORA and Corvis ST are the two noninvasive instruments that are commercially available for corneal biomechanic measurement in clinical settings.[ 21–26 ] In the present study, we used Corvis ST as it evaluates corneal biomechanics from complete dynamic corneal deformation process with superior ability to differentiate FF and subclinical KC from normal eyes.[ 11 ]
Corneal epithelium is not homogenous depth over the Bowman’s layer. Epithelium compensates for stromal surface irregularities, becoming thicker over valley and thinner over hills where stroma bulges. Such epithelium remodeling masks the area of cone in early stages of KC making it difficult to diagnose such modifications on topography. Reinstein et al . have described the doughnut shape of epithelial thinning surrounded by thick epithelium in KC.[ 20 ] Corneal epithelium mapping helps to understand epithelial pathologies such as dry eyes, corneal warpage, and epithelial Bowman’s dystrophy, which gives a false steepening on the topography map mimicking a KC.[ 15 ] However, in the present study, we did not see any role of epithelial mapping in detecting FFKC and SBKC. Thus, epithelial mapping as a sole diagnostic refractive screening tool can miss SBKC and FFKC.
Commercial software of SD-OCT displays SD of epithelial thickness which is also an important parameter to distinguish normal corneas from KC.[ 15 ] SD is defined as SD of all epithelial thicknesses recorded in the central 5 mm of the cornea. The mean of SD of corneal epithelial thickness in normal and KC eyes is 1.55 and 5.92, respectively;[ 27 ] in the present study, we took a cutoff of 3. Higher deviation is usually seen in cases with KC.[ 27 ] In the present study, only 10% of cases of FFKC had SD >3, 40% of cases of SB cases had SD >3, and 60% of cases of KC had SD >3, thus suggesting poor role of SD of epithelial thickness in detecting FFKC and SBKC.
Artificial intelligence has played a major role in differentiating KC, subclinical, or FFKC from normal eyes.[ 1 ] An artificial intelligence-based TBI is a composite index which uses raw data from Pentacam and Corvis ST and is analyzed using the random forest classifier. These integrated data are more decisive in differentiating abnormal corneas from normal. In our study, it helped detect all cases (100%) in KC group in contrast to epithelial mapping (77.77%) and CBI (95.5%) [Figure 1 ].
Figure 1: Distribution of epithelial mapping, CBI, and TBI in diagnosing KC. CBI: Corvis biomechanical index, TBI: Tomographic biomechanical index, KC: Keratoconus
A previous study has also reported that the TBI was the most accurate parameter to distinguish SB and mild KC from normal eyes.[ 11 ] Furthermore, the present analysis revealed that CBI and TBI were able to detect 60% and 90%, respectively, in both FFKC [Figure 2 ] and subclinical KC groups [Figure 3 ] in comparison to epithelial mapping 0% and 40% in FFKC and SBKC, respectively. In our study, TBI also helped diagnose the maximum cases of spectrum of KC. Ambrósio et al . have compared the accuracy of BAD-D and CBI to TBI in detecting ectasia and concluded that TBI has greater accuracy than other techniques.[ 11 ]
Figure 2: Distribution of epithelial mapping, CBI, and TBI in diagnosing FFKC. CBI: Corvis biomechanical index, TBI: Tomographic biomechanical index, FFKC: Forme fruste keratoconus
Figure 3: Distribution of epithelial mapping, CBI, and TBI in diagnosing SBKC. CBI: Corvis biomechanical index, TBI: Tomographic biomechanical index, SBKC: Subclinical keratoconus
Thus, we conclude that the utility of epithelial mapping as a solitary tool is limited in detecting the spectrum of KC, especially SB and FFKC. However, combining it with corneal biomechanical parameters could help improve the efficacy of diagnosis of KC. Artificial intelligence-based TBI helps better detection of FF and subclinical KC.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
Acknowledgment
The authors would like to thank Mr. Ganesh Sakharkar for technical help.
References
1. Santodomingo-Rubido J, Carracedo G, Suzaki A, Villa-Collar C, Vincent SJ, Wolffsohn JS. Keratoconus:An updated review. Cont Lens Anterior Eye 2022;45 101559.
2. Gokhale NS. Epidemiology of keratoconus. Indian J Ophthalmol 2013;61:382–3.
3. Jonas JB, Nangia V, Matin A, Kulkarni M, Bhojwani K. Prevalence and associations of keratoconus in rural Maharashtra in Central India:The central India eye and medical study. Am J Ophthalmol 2009;148:760–5.
4. Kymes SM, Walline JJ, Zadnik K, Sterling J, Gordon MO Collaborative Longitudinal Evaluation of Keratoconus Study Group. Changes in the quality-of-life of people with keratoconus. Am J Ophthalmol 2008;145:611–7.
5. Thulasidas M, Teotia P. Evaluation of corneal topography and tomography in fellow eyes of unilateral keratoconus patients for early detection of subclinical keratoconus. Indian J Ophthalmol 2020;68:2415–20.
6. Vinciguerra R, Ambrósio R Jr, Elsheikh A, Roberts CJ, Lopes B, Morenghi E, et al. Detection of keratoconus with a new biomechanical index. J Refract Surg 2016;32:803–10.
7. Mohammadpour M, Etesami I, Yavari Z, Naderan M, Abdollahinia F, Jabbarvand M. Ocular response analyzer parameters in healthy, keratoconus suspect and manifest keratoconus eyes. Oman J Ophthalmol 2015;8:102–6.
8. Tejwani S, Shetty R, Kurien M, Dinakaran S, Ghosh A, Sinha Roy A. Biomechanics of the cornea evaluated by spectral analysis of waveforms from ocular response analyzer and Corvis-ST. PLoS One 2014;9 e97591.
9. Lopes BT, Roberts CJ, Elsheikh A, Vinciguerra R, Vinciguerra P, Reisdorf S, et al. Repeatability and reproducibility of intraocular pressure and dynamic corneal response parameters assessed by the Corvis ST. J Ophthalmol 2017;2017 8515742.
10. Yang K, Xu L, Fan Q, Zhao D, Ren S. Repeatability and comparison of new Corvis ST parameters in normal and keratoconus eyes. Sci Rep 2019;9 15379.
11. Ambrósio R Jr, Lopes BT, Faria-Correia F, Salomão MQ, Bühren J, Roberts CJ, et al. Integration of scheimpflug-based corneal tomography and biomechanical assessments for enhancing ectasia detection. J Refract Surg 2017;33:434–43.
12. Sinha Roy A, Shetty R, Kummelil MK. Keratoconus:A biomechanical perspective on loss of corneal stiffness. Indian J Ophthalmol 2013;61:392–3.
13. Patil R, Shetty R, Narasimhan R, Patel Y, Khamar P, Pircher M, et al. Mapping of corneal birefringence in thin and asymmetric keratoconus corneas with ultrahigh-resolution polarization-sensitive OCT. J Cataract Refract Surg 2022;48:929–36.
14. Li Y, Chamberlain W, Tan O, Brass R, Weiss JL, Huang D. Subclinical keratoconus detection by pattern analysis of corneal and epithelial thickness maps with optical coherence tomography. J Cataract Refract Surg 2016;42:284–95.
15. Khamar P, Rao K, Wadia K, Dalal R, Grover T, Versaci F, et al. Advanced epithelial mapping for refractive surgery. Indian J Ophthalmol 2020;68:2819–30.
16. Belin MW, Kundu G, Shetty N, Gupta K, Mullick R, Thakur P. ABCD:A new classification for keratoconus. Indian J Ophthalmol 2020;68:2831–4.
17. Sidky MK, Hassanein DH, Eissa SA, Salah YM, Lotfy NM. Prevalence of subclinical keratoconus among pediatric Egyptian population with astigmatism. Clin Ophthalmol 2020;14:905–13.
18. Le Q, Chen Y, Yang Y, Xu J. Measurement of corneal and limbal epithelial thickness by anterior segment optical coherence tomography and
in vivo confocal microscopy. BMC Ophthalmol 2016;16 163.
19. Reinstein DZ, Yap TE, Archer TJ, Gobbe M, Silverman RH. Comparison of corneal epithelial thickness measurement between Fourier-Domain OCT and very high-frequency digital ultrasound. J Refract Surg 2015;31:438–45.
20. Reinstein DZ, Gobbe M, Archer TJ, Silverman RH, Coleman DJ. Epithelial, stromal, and total corneal thickness in keratoconus:Three-dimensional display with Artemis very-high frequency digital ultrasound. J Refract Surg 2010;26:259–71.
21. Sharifipour F, Panahi-Bazaz M, Bidar R, Idani A, Cheraghian B. Age-related variations in corneal biomechanical properties. J Curr Ophthalmol 2016;28:117–22.
22. Nemeth G, Szalai E, Hassan Z, Lipecz A, Flasko Z, Modis L. Corneal biomechanical data and biometric parameters measured with Scheimpflug-based devices on normal corneas. Int J Ophthalmol 2017;10:217–22.
23. Kenia VP, Kenia RV, Pirdankar OH. Association between corneal biomechanical parameters and myopic refractive errors in young Indian individuals. Taiwan J Ophthalmol 2020;10:45–53.
24. He M, Wang W, Ding H, Zhong X. Corneal biomechanical properties in high myopia measured by dynamic scheimpflug imaging technology. Optom Vis Sci 2017;94:1074–80.
25. Yang E, Roberts CJ. A Review of corneal biomechanics after LASIK and SMILE and the current methods of corneal biomechanical analysis. J Clin Exp Ophthalmol 2015;06:1–6.
26. Sedaghat MR, Momeni-Moghaddam H, Azimi A, Fakhimi Z, Ziaei M, Danesh Z, et al. Corneal biomechanical properties in varying severities of myopia. Front Bioeng Biotechnol 2020;8 595330.
27. Catalan S, Cadarso L, Esteves F, Salgado-Borges J, Lopez M, Cadarso C. Assessment of corneal epithelial thickness in asymmetric keratoconic eyes and normal eyes using Fourier domain optical coherence tomography. J Ophthalmol 2016;2016 5697343.
