To determine the impact of corneal epithelial thickness maps on screening for refractive surgery candidacy in a single refractive surgical practice.
Comparison of screening methods.
A total of 100 consecutive patients who presented for refractive surgery screening were evaluated. For each patient, screening based on Scheimpflug tomography, clinical data, and patient history was performed and a decision on eligibility for laser in situ keratomileusis (LASIK), photorefractive keratectomy (PRK), and small incision lenticule extraction (SMILE)was independently made by 2 masked examiners. Examiners were then shown patients’ epithelial thickness maps derived from optical coherence tomography (OCT). The percentage of screenings that changed after evaluating the epithelial thickness maps, with regard to candidacy for surgery, and ranking of surgical procedures from most to least favorable was determined.
Candidacy for corneal refractive surgery changed in 16% of patients after evaluation of the epithelial thickness maps, with 10% of patients screened in and 6% screened out. Surgery of choice changed for 16% of patients, and the ranking of surgical procedures from most to least favorable changed for 25% of patients. A total of 11% of patients gained eligibility for LASIK, whereas 8% lost eligibility for LASIK. No significant difference was found between the evaluations of the 2 examiners.
Epithelial thickness mapping derived from optical coherence tomography imaging of the cornea altered candidacy for corneal refractive surgery, as well as choice of surgery, in a substantial percentage of patients in our practice, and was thus a valuable tool for screening evaluations. Overall, the use of epithelial thickness maps resulted in screening in a slightly larger percentage of patients for corneal refractive surgery.
T he corneal epithelium has a remarkable capacity to remodel in order to preserve regular corneal curvature and thereby maintain visual quality. The epithelial thickness profile of a normal cornea is relatively uniform and may exhibit a certain degree of inferior thickening, whereas in ectatic corneal disorders or other causes of irregular stromal curvature, epithelial thickness will vary widely across the cornea. Regional epithelial thickness mapping has provided insight into the remodeling patterns that occur due to ectatic disease , , and after corneal refractive surgery , and has proven valuable in corneal evaluations. , In ectatic corneal disease, there is a characteristic focal epithelial thinning or “doughnut” pattern that is typically colocalized with focal anterior steepening in manifest cases, , , whereas in the earliest disease manifestations, subtle epithelial thinning can be accompanied by less conclusive changes in anterior curvature. Alternatively, what may appear as suspicious focal steepening on anterior curvature may be attributable to epithelial thickening due to nonpathologic causes rather than an ectatic disorder. , ,
Screening patients for corneal refractive surgery candidacy remains challenging because it is critical to identify subtle ectatic disease before dramatic changes occur. Further, despite multiple proposed machine-derived metrics, screening patients for refractive surgical candidacy remains largely interpretive and therefore subjective because none of these metrics have proven to be equivalent to or better than expert interpretation of corneal images. Placido-based corneal topography and Scheimpflug corneal tomography are currently the most used screening technologies, but neither technology provides epithelial thickness mapping. Both high-frequency digital ultrasound , , , and optical coherence tomography , , , , can provide detailed regional epithelial thickness profiles.
As with all other corneal imaging, epithelial thickness maps require subjective pattern interpretation. In screening for corneal refractive surgery candidacy, Reinstein and associates reported that the addition of epithelial thickness mapping could either rule out patients who are seemingly reasonable candidates based on Placido-based topography or Scheimpflug tomography alone, or rule in candidates with suspicious steepening patterns when these patterns are accompanied by focal epithelial hyperplasia. In our clinical practice, we use epithelial thickness maps adjunctively with Scheimpflug imaging as part of our routine screening practice in conjunction with patient-related factors, including age, corneal thickness, and refractive error to be treated. Similar to prior reports, we have perceived that perceived thatthe use of adjunctive epithelial thickness maps influences our decision-making both in terms of candidacy for surgery and for our preferred surgical technique. However, the extent and manner in which epithelial thickness maps influence surgical candidacy has not, to our knowledge, been directly evaluated.
The purpose of this study was to determine the impact of corneal epithelial thickness mapping on surgical decision-making in a single refractive surgery clinic.
A total of 100 consecutive patients who presented for refractive surgery screening at the Cole Eye Institute between February 2020 and September of 2020 were independently evaluated by 2 examiners (J.B.R. and W.J.D.) in a masked fashion at least 6 months after their initial clinical evaluation. To further reduce the risk of recall bias, examiners were presented with cases drawn from both practices and without knowledge of which practice each case came from. The study adhered to the tenets of the Declaration of Helsinki. Only deidentified patient data were used; therefore, the study was determined by the institutional review board of the Cleveland Clinic to be exempt from requiring explicit approval.
Patients were included if they had undergone a comprehensive ophthalmic examination at the time of their clinical evaluation, including a slitlamp examination, dilated fundus examination, and determination of their manifest and cycloplegic refractions, as well as imaging with a Scheimpflug tomographer (Pentacam HR version 6.09r43; Oculus Optikgeräte GmbH) and an anterior segment optical coherence tomographer (AS-OCT) (Avanti RTVue XR version 2018.1.1.63; Optovue). Patients were excluded if they had imaging of inadequate quality, had a history of previous ocular surgery, or had any ophthalmic abnormality that would alter surgical decision-making regardless of their risk of corneal ectasia (eg, corneal scars). Soft contact lens wear was discontinued at least 3 days before initial screening. None of the patients who presented wore hard contact lenses.
An initial evaluation was conducted by each examiner assessing a secure repository of deidentified screenshots from the electronic medical record (Epic EMR; Epic Systems Corporation) including (1) patients’ age, (2) pertinent nonidentifying medical and ophthalmic history as documented by the technician at the initial refractive surgery screening visit, (3) textual ophthalmic examination findings (manifest and cycloplegic refractions, slitlamp examination, and posterior segment examination), (4) the 4-map refractive display from the Scheimpflug tomographer displaying axial topography, pachymetry, anterior elevation referenced to best fit sphere, and posterior elevation referenced to best fit sphere, and (5) the Belin/Ambrósio Enhanced Ectasia Display from the Scheimpflug tomographer. Patients were evaluated for candidacy for corneal refractive surgery (laser in situ keratomileusis [LASIK], photorefractive keratectomy [PRK], and small incision lenticule extraction [SMILE]) based on a collective assessment of both eyes for each patient. For patients who were deemed suitable candidates for surgery, the surgical procedures were ranked from most favorable to least favorable. All patients included in the analysis were otherwise candidates for refractive surgery based on refractive error and other factors; thus, candidacy for surgery and for specific procedures was decided based on risk of ectasia, with LASIK considered the most biomechanically destabilizing of the procedures, SMILE being equivalent to LASIK, or potentially offering a slight biomechanical advantage over LASIK, and PRK considered the least destabilizing of the three and a reasonable option for patients who present with mildly suspicious findings that were not extensive enough to preclude them from undergoing a corneal refractive procedure.
The examiners were then shown the patients’ epithelial thickness maps provided by the AS-OCT and were asked to reconsider their evaluation with all information available. An epithelial thickness pattern finding of central or inferior thinning increased the examiners’ concern about patients’ risk for ectasia. Depending on the extent and pattern of thinning, patients who had been deemed candidates for surgery on initial re-evaluation either lost eligibility for corneal refractive surgery, remained candidates for a more conservative procedure (PRK) but lost eligibility for more biomechanically impactful procedures (eg, LASIK), had a change in the ranking of their surgical options, or did not have any change in their evaluation. Conversely, an epithelial thickness pattern finding of focal thickening, particularly when colocated with mildly suspicious focal steepening on anterior curvature, decreased the examiners’ concern for ectasia. Depending on the extent and pattern of thickening, and its correlation with anterior curvature findings, patients who were deemed poor candidates for surgery on initial re-evaluation either remained poor candidates for surgery or gained eligibility for corneal refractive surgery. In addition, patients with such findings who were initially deemed candidates for surgery either did not have any change in their evaluation, had a change in the ranking of their surgical options, or gained eligibility for the more biomechanically weakening procedures (SMILE and/or LASIK) if they were initially only eligible for PRK.
Based on these general approaches to interpretation, the percentage of evaluations that resulted in a change in candidacy for surgery, as well as the percentage of patients for whom the surgery of choice and the ranking of surgical procedures changed, was determined. In the subset of patients whose evaluations, after assessment of their epithelial thickness maps, did not change in terms of candidacy for surgery and ranking of surgical procedures from most to least favorable, the examiners noted whether the epithelial thickness maps increased, decreased, or had no effect on their confidence in their decision.
Patient demographics for the 100 patients evaluated in this study are shown in Table 1 . Overall, 82 patients (82%) were determined to be candidates for corneal refractive surgery, whereas 18 patients (18%) were deemed to be unsatisfactory candidates and excluded from corneal refractive surgery.
|Mean ± Standard Deviation (Range)
|33 ± 9 (18 to 60)
|53% females, 47% males
|Manifest sphere (D)
|−4.35 ± 2.88 (5.50 to −12.25)
|Manifest cylinder (D)
|0.79 ± 0.85 (0.00 to 4.25)
|Manifest spherical equivalent (D)
|−3.95 ± 2.88 (6.50 to −11.50)
|Corrected distance visual acuity (logMAR)
|−0.08 ± 0.07 (0.20 to −0.20)
|44.93 ± 1.52 (40.70 to 49.80)
|Thinnest pachymetry (μm)
|544.19 ± 32.70 (457.00 to 617.00)
|0.98 ± 0.65 (−0.68 to 3.54)
Candidacy for corneal refractive surgery changed in 16% of patients after evaluation of their epithelial thickness maps, with 10% of patients screened in and 6% screened out ( Figure 1 ). In other words, 9 to 10 patients who were initially deemed unsuitable for corneal refractive surgery based on Scheimpflug tomography alone were ultimately offered surgery when epithelial thickness maps were also considered, whereas 5 to 7 patients who were initially considered reasonable surgical candidates were ultimately not offered surgery, having been ruled out based on epithelial thickness maps. Figure 2 shows a case example where the decision was made to offer surgery when it was initially declined, whereas Figure 3 shows a case example where the decision was made to decline corneal refractive surgery when surgery was initially deemed a reasonable option. Figure 4 shows a case example where surgical candidacy was not affected by epithelial thickness map findings.