An increasing percentage of the estimated one million people in the United States who undergo refractive surgery annually, are hyperopic. To date, most topographic pre-surgical screening normal values have been based on a predominantly myopic population [1,2]. This is mainly due to the fact that, in the past, the vast majority of individuals undergoing refractive surgery were myopic, and published normal values often reflected the propensity of myopes in the early study populations. It was unknown; however, whether these myopic biased values were applicable to a hyperopic population.It is well-established that the myopic and hyperopic populations are diverse. Hyperopic eyes are associated with a shallower anterior chamber [3], increased incidence of narrow angle glaucoma [4], shorter axial length, higher sphericity, and higher total and corneal spherical aberration compared to myopic eyes [5]. Normative data bases that are not applicable to the subset being examined can lead to both false positives and false negatives. Specific population-based normal values can have important clinical implications.

Prior clinical observations have led us to believe that there is an increased variability in the posterior elevation in hyperopic eyes on tomographic evaluation. If this is true, it would lead to false positives when compared against a myopic biased normative database, and potentially exclude "normal" patients from consideration for refractive surgery. The goals of our study are to establish a hyperopic normative data base for elevation-based tomography and to compare the data to a previously studied myopic group.

Institutional Review Board approval was obtained prior to beginning the study. We conducted a retrospective review of data from 100 normal hyperopic eyes of 51 patients who presented for refractive surgery evaluation. All eyes were hyperopic. Mixed astigmatic eyes were excluded, regardless of the spherical equivalent. Anterior and posterior corneal elevation measurements, as well as corneal pachymetry measurements at the apex and thinnest points were collected using a rotating Scheimplugtomographer, PENTACAM HR (OCULUS GmbH, Wetzler, Gemany), according to a previously published protocol. [6]

Calibration of the PENTACAM was performed by the manufacturer using a model eye and this was not repeated immediately prior to data collection. Images were obtained by technicians who had extensive prior experience with the PENTACAM. Contact lens use was discontinued at least two weeks prior to corneal imaging. Patients with previous ocular surgery or known ectasia were excluded from the study, as well as any patient with a central pachymetry reading below 475 µm or greater than 650 µm, best spectacle-corrected visual acuity (BSCVA) below 20/20, visible slit lamp abnormalities, or a family history of keratoconus. Patients were asked to blink twice and then look at the PENTACAM fixation device. Image acquisition involved a 1-second scan of 25 rotational Scheimpflug images. Acceptable maps had at least 9.0 mm of corneal coverage with no extrapolated data in the central 8.0 mm zone. When the patient blinked during the scan or other artifacts were introduced, the PENTACAM flagged that scan as not acceptable. Scans not meeting acceptable criteria were repeated.

Anterior and posterior corneal elevation measurements and corneal pachymetry measurements were documented at the corneal apex and thinnest point. The elevation data used to calculate the best fit sphere (BFS) was gathered from a fixed 8.0 mm diameter zone centered on the apex. There was no evidence of extrapolated data in this 8.0 mm zone and no extrapolated data was used to calculate the BFS. Corneal elevation was measured off of a BFS as calculated by the PENTACAM. The mean, range and standard deviation for elevation at the apex and thinnest point of the anterior and posterior corneal surfacewas determined.The Kolmogorov-Smirnov test was used to evaluate for parametric data distribution. Both the Student T-Test (parametric data distribution) and Mann-Whitney Test (non-parametric distribution) were utilized to determine statistical differences between the two populations.

A total of 100 hyperopic eyes of 51 patients were evaluated to determine normal values for corneal pachymetry as well as anterior and posterior elevation. The mean patient age was 53.5 ± 8.7 years (SD), with a range of 27 to 68years. There were 29 men and 22 women included in the study. The average pachymetry was 550 ± 33.0 µm at the apex (range 484-626 µm) and 545 ± 33.2 µm at the thinnest point (range 479-626 µm). Average anterior elevation values at the apex and thinnest points were 0.4 µm and -0.1 µm, respectively. Average posterior elevation values at the apex and thinnest points were 5.7 µm and 10.6 µm, respectively. The range and standard deviations of the elevation values are shown in Table 1.

Table 1: Average, range and standard deviation of normal hyperopic corneal elevation values.

An earlier studied myopic group from the same refractive practice included a total of 100 eyes of 50 patients [5]. The mean patient age was 40.1 ± 10.1 years (SD), with a range of 18 to 60 years. There were 24 men and 26 women included in the study. The average pachymetry of these patients was 550 ± 36.1 µm at the apex (range 453-661 µm) and 548± 36.3 µm at the thinnest point (range 446-660 µm). Average anterior elevation values at the apex and thinnest points were 1.6 µm and 1.7 µm respectively. Average posterior elevation values at the apex and thinnest points were 0.8 µm and 3.6 µm respectively. The range and standard deviations of the elevation values for the myopic group are shown in Table 2.

Table 2: Average, range and standard deviation of normal myopic corneal elevation values.

The Kolmogorov-Smirnov test confirmed a normal data set for all the measurement groups, except the anterior elevation at the apex for hyperopes, and the anterior elevation at the apex and thinnest point in myopes. The results of the Student T-Test and Mann-Whitney Test are shown in Table 3.

Table 3: Statistical comparison for all hyperopic and myopic eyes.

Increasing patient and professional expectations regarding refractive surgery requires us to continually refine our screening process in order to minimize avoidable postsurgical complications. Tomographic evaluation is becoming common practice for the presurgical evaluation of refractive surgery candidates. It creates a three dimensional reconstruction of the anterior segment by measuring the anterior and posterior corneal surfaces, as well as the anterior and posterior lens and iris. Refractive screening concentrates more on the corneal surfaces and their spatial relationship (i.e. corneal thickness). The lack of standardized normal elevation values specific for the hyperopic population has likely led to false positives when these patients are screened against the currently used myopic biased measurements. We studied a hyperopic population in order to establish a normative database and to highlight any significant differences between myopic and hyperopic populations.

Statistical analysis detected significant differences between measurements in the hyperopic population compared to the myopic population. The differences in anterior elevation at the apex (P=6.38x10^-7) and at the thinnest point (P=1.09x10^-8) are highly significant. The differences in posterior elevation at the apex (4.88x10^-21) and at the thinnest point (P=1.07x10^-17) are of even greater significance. This confirms our observational hypothesis that there appears to be innate corneal elevation differences between the two populations.

Given that each eye is not a totally independent variable, analysis was carried out on each eye individually in addition to both eyescombined. [7] Similar significance was obtained in all groups. The right eye anterior elevation measurements at the apex and thinnest point were significantly different (P=0.0076 and P=0.00011, respectively), as well as the difference in posterior elevation at the apex and thinnest point (P=2.23x10^-12 and P=1.14x10^-10 respectively).Analysis of the left eye showed similar results.

The demographics of the two populations (hyperopic and myopic) are compared in Table 4.The observed elevation differences between the myopic and hyperopic groups may have, in part, been influenced by the age variance between the two study groups [8]. Therefore, we analyzed an age-matched group from the myopic population (n=36) to within 1 SD of the average hyperopic population (n=75), which included those individuals between the ages of 45-62. The previously noted pachymetric values remained similar between age-matched groups. The changes in anterior elevation lost significance in the age-matched pairing, while the changes in posterior elevation remained highly significant (Table 5). The loss of significance for the anterior elevation values may be in part due to the smaller sample size, but likely has little clinical significance (see below).

Table 4: Demographic information comparing the myopic and hyperopic populations.

Table 5: Age-matched subgroup analysis, when comparing all myopic and hyperopic eyes aged 45-62.

While the posterior results substantiated our earlier clinical observations, the initial anterior results were unexpected. Although the differences in anterior measurements reached statistical significance, they are unlikely to appreciably alter our screening parameters as the +2 SD and +3 SD values, often used as screening gates, between the myopic and hyperopic groups are similar (within 1.4 µm). The typically used screening gates (+2 SD/+3 SD) for the posterior surface would vary greatly as they differ by 6.1 and 6.8 µm for the apex and 9.1 and 10.1 µm for the thinnest point. These differences must be accounted for when screening hyperopic refractive surgical candidates, and this highlights the need to define a normative database specific for a hyperopic population.

In this study we computed the BFS by fixing the area utilized to the central 8.0 mm zone. No extrapolated data was used to calculate the BFS as incorporation of extrapolated data may lead to inconsistencies in the calculation of the BFS. All of our previous studies have utilized an 8.0 mm zone and this is also the zone size utilized by the manufacturer for some of their fixed displays [6,9,10]

It should be noted that data was measured at the corneal apex and thinnest points only. These are two points that can be easily and consistently identified when looking at multiple elevation maps from different patients. Only elevation measurements at these two points can be compared to the normative data presented in this study. There may be significant areas of elevation in the peripheral cornea as a function of normal astigmatism. An accurate assessment of image quality must also be made.

While the PENTACAM has been shown to have excellent agreement with the ultrasound pachymeter with regard to central corneal thickness in both pre and post LASIK eyes [11], there are considerable variability in elevation values across different elevation based systems12. This potentially limits the normative data developed in this study to the PENTACAM Eye Scanner.

Our results suggest that anterior elevations greater than 4.2µm at the apex or 4.3µm at the thinnest point occur in less than 5% of normal hyperopic corneas (2SD). Anterior elevations greater than 6.1µm at the apex or 6.5µm the thinnest point occur in less than 0.3% of normal corneas (3SD). Posterior elevations of greater than 12.9µm at the apex and 22.1µm at the thinnest point occur in less than 5% of normal corneas (2SD). Posterior elevations of greater than 16.6µm at the apex and 27.8µm at the thinnest point occur in less than 0.3% of normal corneas (3SD). Anterior or posterior elevation values greater than 2 to 3 standard deviations above the mean should raise suspicion and may warrant further evaluation. The posterior values at both the apex and thinnest point are appreciably dissimilar from myopic values to warrant separate screening parameters for hyperopic patients.

Although our patient population was limited in size, and further reduced in the age matched comparison subgroup, there was sufficient diversity to maintain statistical significance for the posterior elevation data (Table 5). Future studies with a larger population would help validate this initial study. While a larger cohort may ultimately reach statistical significance for the anterior surface, the changes are not likely to be clinically relevant as the differences were small. Furthermore, while single topographic measurements may be helpful for screening purposes; no single parameter can reliably define a multi-factorial condition such as keratoconus. Other means of examining the topographic maps including pattern recognition, OD/OS comparison, and pachymetry should be utilized [13,14].

Drs. Belin and Ambrosio serve as Consultants to Oculus GmbH (Wetzlar, Germany). The other authors disclose no commercial or financial interest.