- Email: [email protected]
Comparison of Pentacam and Orbscan IIz on Posterior Curvature Topography Measurements in Keratoconus Eyes
Comparison of Pentacam and Orbscan IIz on Posterior Curvature Topography Measurements in Keratoconus Eyes Susannah Quisling, MD,1 Stacy Sjoberg, MD,1 Bridget Zimmerman, PhD,2 Kenneth Goins, MD,1 John Sutphin, MD1 Purpose: To compare the measurements of posterior elevation above the best-fit sphere and of pachymetry in established keratoconus patients using 2 different technologies—Scheimpflug imaging with the Pentacam and scanning slit combined with Placido imaging with Orbscan IIz. Design: Retrospective comparative case series. Participants: Patients with confirmed keratoconus by biomicroscopy and Placido topography who had corneal topography scans by the Pentacam and Orbscan IIz on their visit from April 2004 to April 2005. These scans were obtained routinely for patient diagnosis, treatment, and progression. Methods: The mean difference, standard deviation (SD), and 95% limits of agreement (LOA) were calculated, and Bland–Altman plots were constructed for best-fit sphere radius of curvature, posterior elevation above the best-fit sphere, and pachymetry of the thinnest point. Results: Thirty-six eyes of 29 patients were analyzed. The average radius for posterior best-fit sphere was 5.97 m (range, 4.69 – 6.79) for the Pentacam and 6.00 m (range, 4.97– 6.55) for Orbscan IIz. The mean difference (Pentacam ⫺ Orbscan IIz) for the radius generated for the best-fit sphere was ⫺0.03⫾0.22 SD with a 95% confidence interval (CI) of ⫺0.11 to 0.04 and a 95% LOA of ⫺0.46 to 0.40 (P ⫽ 0.362). The mean posterior elevation by best-fit sphere fixed to the apex was 34.86 m (range, 3–120) for the Pentacam and 48.50 m (range, 11–118) for Orbscan IIz. The mean difference was ⫺13.64⫾26.08 SD (95% CI, ⫺22.46 to ⫺4.81; 95% LOA, ⫺64.75 to 37.48; P ⫽ 0.003). The average thinnest point for the Pentacam was 443.6 m (range, 164 –587), and that for Orbscan IIz was 445.9 m (range, 204 –597). The mean difference in the thinnest point for the Pentacam and Orbscan IIz was ⫺2.28⫾35.55 (95% CI, ⫺14.31 to 9.75; 95% LOA, ⫺71.95 to 67.39; P ⫽ 0.703). Conclusion: Both the Pentacam and Orbscan IIz determine similar thinnest points but have a measurable difference in posterior elevations above the best-fit sphere, despite similar radii of curvature. This difference may be important in the screening of patients for refractive surgery to avoid surgery on patients with early keratoconus. This study cannot determine if the Pentacam is underestimating the posterior vault or if Orbscan is overestimating this height, and further studies with a standardized test object are needed. Ophthalmology 2006; 113:1629 –1632 © 2006 by the American Academy of Ophthalmology.
The characteristic stromal thinning in keratoconus leads to corneal topography changes on Pentacam (Oculus, Inc., Lynnwood, WA) and Orbscan IIz (Bausch & Lomb, Rochester, NY) topography. This study was designed to compare the measurements of the radius of curvature, posterior elevation above the best-fit sphere, and pachymetry in a popOriginally received: November 30, 2005. Accepted: March 7, 2006. Manuscript no. 2005-1173. 1 Department of Ophthalmology and Visual Sciences, University of Iowa Hospitals and Clinics, Iowa City, Iowa. 2 Department of Biostatistics, University of Iowa, Iowa City, Iowa. Presented as a poster at: American Academy of Ophthalmology meeting, October 2005, Chicago, Illinois. Supported by the University of Iowa Department of Ophthalmology Resident and Fellow Research Fund. Reprint requests to John Sutphin, MD, University of Iowa Ophthalmology and Visual Sciences, 200 Hawkins Drive, 11290 PFP, Iowa City, IA 52242-1091. © 2006 by the American Academy of Ophthalmology Published by Elsevier Inc.
ulation of keratoconus patients who are noted to have high posterior curves and to validate the measurements by using 2 different imaging methodologies. Orbscan IIz has been commercially available since 19991 and is based on a slit-scanning beam projected on the cornea and Placido image acquisition. The Pentacam has been available since 2004 and is based on a capture method called Scheimpflug imagery, which measures 25 000 elevation points and gives a representation of the shape of the corneal surface. There is much in the literature describing Orbscan IIz and its use, but there is little in the literature about the Pentacam and its use. Until the Pentacam became commercially available, there had been no method of comparing posterior cornea measurements and determining their accuracy. If these 2 machines find accurate measurements, then the data produced by both machines with regard to the posterior curvature maps are validated. ISSN 0161-6420/06/$–see front matter doi:10.1016/j.ophtha.2006.03.046
1629
Ophthalmology Volume 113, Number 9, September 2006 puted. Bland–Altman plots show the plot of the observed differences, with the mean bias and 95% LOA constructed.
1.20
0.40
0.40
0.00
-0.034
-0.40
-0.46
-0.80
-1.20 4.60
5.00
5.40
5.80
6.20
6.60
7.00
Radius of best fit sphere (average of Pentacam and Orbscan)
Figure 1. Bland–Altman plot showing the mean difference of the radii generated for the best-fit sphere by the Pentacam and Orbscan IIz (⫺0.03⫾0.22 m [standard deviation]; P ⫽ 0.362; 95% confidence interval, ⫺0.11 to 0.04). This graph shows that the 95% limits of agreement were between ⫺0.46 and 0.40, indicating that most of the difference in the radius was no greater than ⫺0.46 and ⫹0.40.
Materials and Methods The Pentacam and Orbscan IIz databases were searched retrospectively from April 2004 to June 2005 for all patients with the diagnosis of keratoconus confirmed by either of the two cornea specialists at the University of Iowa Hospital and Clinics, Iowa City, Iowa. Twenty-nine patients with the diagnosis of keratoconus who had corneal topography from both systems were identified. Only patients who had both scans ordered and performed for diagnosis, treatment, and progression were included. Data were collected from the Pentacam Comprehensive Eye Scanner (version 1.09) and Orbscan IIz topography system (version 3.2). There were a total of 36 scans used from the 29 patients, as 15 eyes had undergone corneal transplant; 2 eyes had significant lid artifact from blinking, 1 patient had only one scan and was missing an Orbscan IIz scan, and 4 eyes were not diagnosed with keratoconus. Pentacam and Orbscan IIz images were obtained on the same day. To obtain the most similar data for comparison, the published literature was reviewed. For Orbscan IIz, posterior elevation was measured by the best-fit sphere fixed to the apex with full corneal fit over a zone diameter of 9 m. For the Pentacam, the diameter was controlled at 9 m and the float option was removed, giving the best-fit sphere for that scan of the cornea. The radius of curvature that both measured was compared, as well as pachymetry and position of the thinnest point. Agreement of the various measurements obtained using the Pentacam and Orbscan IIz was assessed using the method described by Bland and Altman.2 An estimate of the mean bias, as measured by the mean of the paired differences between the 2 methods, was obtained. The 1-sample t test was performed to test if the mean bias significantly differed from zero. The 95% limits of agreement (LOA) (mean difference ⫾ 1.96 standard deviation [SD]), which define the range within which most differences between measurements by the 2 methods will lie, also were com-
1630
Results The patient population consisted of 13 females and 16 males with the diagnosis of keratoconus. The average age was 46 years (range, 23–78). Thirty-five of the cones were located in the inferior temporal quadrant and 1 in the inferior nasal quadrant. The average of the radius generated for the best-fit sphere for the Pentacam was 5.97 m (range, 4.69 – 6.79), and that for Orbscan IIz was 6.00 m (range, 4.97– 6.55). Figure 1 shows the Bland–Altman plot for the mean difference of the radius of the best-fit sphere, which was ⫺0.03⫾0.22 SD. The 95% confidence interval (CI) of the mean difference was ⫺0.11 to 0.04. Statistical comparison of the mean difference of the radius generated by the best-fit sphere fixed to the apex was not found to be significant (P ⫽ 0.362; 95% LOA, ⫺0.46 to 0.40). Means for the posterior elevation by best-fit sphere fixed to the apex were 34.86 m (range, 3–120) by the Pentacam and 48.50 m (range, 11–118) by Orbscan IIz. In Figure 2, there is a Bland– Altman plot of the mean difference of the highest elevation measured, which was ⫺13.64⫾26.08 SD (95% CI, ⫺22.46 to ⫺4.81; 95% LOA, ⫺64.75 to 37.48; P ⫽ 0.003). Average thinnest points were 443.6 m (range, 164 –587) for the Pentacam data and 445.9 m (range, 204 –597) for Orbscan IIz data. The mean difference in the thinnest point for the Pentacam and Orbscan IIz was ⫺2.28⫾35.55 (95% CI, ⫺14.31 to 9.75; 95% LOA, ⫺71.95 to 67.39; P ⫽ 0.703). The Bland–Altman plot for the mean difference is shown in Figure 3.
60 40 Highest elevation difference (Pentacam - Orbscan)
Radius difference (Pentacam - Orbscan)
0.80
37.48
20 0 -13.64
-20 -40 -60
-64.76
-80 -100 -120 -140 0
20
40
60
80
100
120
140
Highest elevation (average of Pentacam and Orbscan)
Figure 2. Bland–Altman plot of the mean difference of the elevation of the posterior curve generated by the Pentacam and Orbscan IIz. The mean difference of the highest elevation obtained was ⫺13.64⫾26.08 m (standard deviation) (95% confidence interval, ⫺22.46 to ⫺4.81; P ⫽ 0.003). The graph shows the 95% limits of agreement to be between ⫺64.76 and 37.48, indicating that there was a greater proportion of larger measurements by Orbscan IIz.
Quisling et al 䡠 Posterior Curvature Topography Measurements in Keratoconus Eyes 240
Thinnest point difference (Pentacam - Orbscan)
160
80
67.39
0
-2.28
-71.95
-80
-160
-240 160
240
320
400
480
560
640
Thinnest point (average of Pentacam and Orbscan)
Figure 3. Bland–Altman plot showing the difference between the thinnest point found by the Pentacam and Orbscan IIz. The mean difference in the thinnest point was ⫺2.28⫾35.55 m (95% confidence interval, ⫺14.31 to 9.75; P ⫽ 0.703). This graph shows that most of the difference between the measurements is no greater than ⫺71.95 and 67.39 (95% limits of agreement).
Discussion Corneal topography often aids in the diagnosis of keratoconus. The characteristic stromal thinning in keratoconus corresponds to an increased posterior elevation above the best-fit sphere on corneal topography. The Pentacam Comprehensive Eye Scanner is based on Scheimpflug technology, and the Orbscan is based on a scanning-slit beam with a Placido disk. Because Orbscan derives the posterior elevation map mathematically, there had been caution raised regarding the posterior curve in the past literature. This retrospective study found a statistically significant difference between posterior elevations above the best-fit sphere that Orbscan IIz and the Pentacam generated, despite no difference in the radii of curvature of the sphere. This difference may be related to the way the 2 machines obtain or interpret their data. Orbscan IIz uses slit scanning/Placido ring technology to measure the cornea surface. The first Orbscan topographer was based on a slit-scanning beam projected on the cornea.1 Then, in 1999 Orbscan IIz added a Placido image to obtain curvature measurements directly. Data points from the reflection of the mires and the 20 slit scans are used to make a representation of the anterior surface of the cornea. Orbscan IIz digitally recreates the posterior cornea, anterior iris, and lens from the collected data as previously described.1 This recreation requires sophisticated triangulation of the previously generated elevation and curvature of the anterior topography.1,3,4 Unlike Orbscan IIz, the Pentacam is based on a capture method called Scheimpflug imagery. By measuring 25 000 elevation points, it gives a representation of the shape of the
corneal surface. These elevation points are independent of axis, orientation, and position and yield only one possible curvature map. The difference in the calculated radii for the posterior best-fit sphere of the Pentacam and Orbscan IIz was not found to be statistically significant, which makes the difference in the measured posterior elevation between the machines important, as this could point to a difference in data analysis by the machines. Analogous to the anterior curve, Orbscan IIz may be more accurate in the periphery but less so in the center, thereby overestimating the posterior height. In the study, Orbscan IIz seems to generate an elevation above the best sphere that is greater than that of the Pentacam. Because most of the literature is based on Orbscan IIz, interpretation of the Pentacam should be based on individual experience with the machine. In our study, the pachymetry of the thinnest point calculated by the Pentacam and Orbscan IIz showed no significant difference in mean deviation. Orbscan has been shown to yield values for pachymetry similar to those of ultrasound in normal corneas and in keratoconus eyes.5,6 Both machines similarly found the location of the cones to be primarily in the inferior temporal quadrant, as in previous studies.7–10 There can be various reasons why there was a difference in the posterior elevation. Tear film abnormalities and the presence of anterior stromal scarring may contribute to this difference. In the small selection of patients who had both the scans present and diagnosis of keratoconus, some of the most disparate comparisons were in patients who had documented contact lens use. More studies will have to be performed in the future to elucidate this more clearly— possibly with a standardized test object. Both the Pentacam and Orbscan IIz provide useful information with regard to location of the keratoconus cone and pachymetry, despite obtaining the topography by different mechanisms. From these data, it appears that Orbscan determines the posterior elevation above the best-fit sphere higher than the Pentacam. This difference is probably due to the difference in methods between the 2 systems, as the scanning-slit topographer estimates the central 3 m, whereas the Scheimpflug images the cornea directly. But there are other mechanisms that could explain the disparity, including density of the pixels, the software used to image the cornea, and clarity of the cornea or tear film. The overall reason for the difference has yet to be elucidated, and more studies are needed. The importance of the study is to understand that when using the machines a direct comparison may not be possible, but the surgeon and clinician should become adept at interpreting the machine they use routinely.
References 1. Cairns G, McGhee CN. Orbscan computerized topography: attributes, applications, and limitations. J Cataract Refract Surg 2005;31:205–20. 2. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135– 60.
1631
Ophthalmology Volume 113, Number 9, September 2006 3. Snook RK. Pachymetry and true topography using the Orbscan system. In: Gills JP, Sanders DR, Thornton SP, Martin RG, eds. Corneal Topography: The State of the Art. Thorofare, NJ: SLACK Inc.; 1995:89 –103. 4. Cairns G, Collins A, McGhee CN. A corneal model for slit-scanning elevation topography. Ophthalmic Physiol Opt 2003;23;193–204. 5. Modis L Jr, Langenbucher A, Seitz B. Evaluation of normal corneas using the scanning-slit topography/pachymetry system. Cornea 2004;23:689 –94. 6. Gherghel D, Hosking SL, Mantry S, et al. Corneal pachymetry
1632
7. 8. 9. 10.
in normal and keratoconic eyes: Orbscan II versus ultrasound. J Cataract Refract Surg 2004;30:1272–7. Auffarth GU, Wang L, Volcker HE. Keratoconus evaluation using the Orbscan topography system. J Cataract Refract Surg 2000;26:222– 8. Maquire LJ, Bourne WM. Corneal topography of early keratoconus. Am J Ophthalmol 1989;108:107–12. Wilson SE, Klyce SD. Advances in the analysis of corneal topography. Surv Ophthalmol 1991;35:269 –77. Wilson SE, Lin DT, Klyce SD. Corneal topography of keratoconus. Cornea 1991;10:2– 8.
The characteristic stromal thinning in keratoconus leads to corneal topography changes on Pentacam (Oculus, Inc., Lynnwood, WA) and Orbscan IIz (Bausch & Lomb, Rochester, NY) topography. This study was designed to compare the measurements of the radius of curvature, posterior elevation above the best-fit sphere, and pachymetry in a popOriginally received: November 30, 2005. Accepted: March 7, 2006. Manuscript no. 2005-1173. 1 Department of Ophthalmology and Visual Sciences, University of Iowa Hospitals and Clinics, Iowa City, Iowa. 2 Department of Biostatistics, University of Iowa, Iowa City, Iowa. Presented as a poster at: American Academy of Ophthalmology meeting, October 2005, Chicago, Illinois. Supported by the University of Iowa Department of Ophthalmology Resident and Fellow Research Fund. Reprint requests to John Sutphin, MD, University of Iowa Ophthalmology and Visual Sciences, 200 Hawkins Drive, 11290 PFP, Iowa City, IA 52242-1091. © 2006 by the American Academy of Ophthalmology Published by Elsevier Inc.
ulation of keratoconus patients who are noted to have high posterior curves and to validate the measurements by using 2 different imaging methodologies. Orbscan IIz has been commercially available since 19991 and is based on a slit-scanning beam projected on the cornea and Placido image acquisition. The Pentacam has been available since 2004 and is based on a capture method called Scheimpflug imagery, which measures 25 000 elevation points and gives a representation of the shape of the corneal surface. There is much in the literature describing Orbscan IIz and its use, but there is little in the literature about the Pentacam and its use. Until the Pentacam became commercially available, there had been no method of comparing posterior cornea measurements and determining their accuracy. If these 2 machines find accurate measurements, then the data produced by both machines with regard to the posterior curvature maps are validated. ISSN 0161-6420/06/$–see front matter doi:10.1016/j.ophtha.2006.03.046
1629
Ophthalmology Volume 113, Number 9, September 2006 puted. Bland–Altman plots show the plot of the observed differences, with the mean bias and 95% LOA constructed.
1.20
0.40
0.40
0.00
-0.034
-0.40
-0.46
-0.80
-1.20 4.60
5.00
5.40
5.80
6.20
6.60
7.00
Radius of best fit sphere (average of Pentacam and Orbscan)
Figure 1. Bland–Altman plot showing the mean difference of the radii generated for the best-fit sphere by the Pentacam and Orbscan IIz (⫺0.03⫾0.22 m [standard deviation]; P ⫽ 0.362; 95% confidence interval, ⫺0.11 to 0.04). This graph shows that the 95% limits of agreement were between ⫺0.46 and 0.40, indicating that most of the difference in the radius was no greater than ⫺0.46 and ⫹0.40.
Materials and Methods The Pentacam and Orbscan IIz databases were searched retrospectively from April 2004 to June 2005 for all patients with the diagnosis of keratoconus confirmed by either of the two cornea specialists at the University of Iowa Hospital and Clinics, Iowa City, Iowa. Twenty-nine patients with the diagnosis of keratoconus who had corneal topography from both systems were identified. Only patients who had both scans ordered and performed for diagnosis, treatment, and progression were included. Data were collected from the Pentacam Comprehensive Eye Scanner (version 1.09) and Orbscan IIz topography system (version 3.2). There were a total of 36 scans used from the 29 patients, as 15 eyes had undergone corneal transplant; 2 eyes had significant lid artifact from blinking, 1 patient had only one scan and was missing an Orbscan IIz scan, and 4 eyes were not diagnosed with keratoconus. Pentacam and Orbscan IIz images were obtained on the same day. To obtain the most similar data for comparison, the published literature was reviewed. For Orbscan IIz, posterior elevation was measured by the best-fit sphere fixed to the apex with full corneal fit over a zone diameter of 9 m. For the Pentacam, the diameter was controlled at 9 m and the float option was removed, giving the best-fit sphere for that scan of the cornea. The radius of curvature that both measured was compared, as well as pachymetry and position of the thinnest point. Agreement of the various measurements obtained using the Pentacam and Orbscan IIz was assessed using the method described by Bland and Altman.2 An estimate of the mean bias, as measured by the mean of the paired differences between the 2 methods, was obtained. The 1-sample t test was performed to test if the mean bias significantly differed from zero. The 95% limits of agreement (LOA) (mean difference ⫾ 1.96 standard deviation [SD]), which define the range within which most differences between measurements by the 2 methods will lie, also were com-
1630
Results The patient population consisted of 13 females and 16 males with the diagnosis of keratoconus. The average age was 46 years (range, 23–78). Thirty-five of the cones were located in the inferior temporal quadrant and 1 in the inferior nasal quadrant. The average of the radius generated for the best-fit sphere for the Pentacam was 5.97 m (range, 4.69 – 6.79), and that for Orbscan IIz was 6.00 m (range, 4.97– 6.55). Figure 1 shows the Bland–Altman plot for the mean difference of the radius of the best-fit sphere, which was ⫺0.03⫾0.22 SD. The 95% confidence interval (CI) of the mean difference was ⫺0.11 to 0.04. Statistical comparison of the mean difference of the radius generated by the best-fit sphere fixed to the apex was not found to be significant (P ⫽ 0.362; 95% LOA, ⫺0.46 to 0.40). Means for the posterior elevation by best-fit sphere fixed to the apex were 34.86 m (range, 3–120) by the Pentacam and 48.50 m (range, 11–118) by Orbscan IIz. In Figure 2, there is a Bland– Altman plot of the mean difference of the highest elevation measured, which was ⫺13.64⫾26.08 SD (95% CI, ⫺22.46 to ⫺4.81; 95% LOA, ⫺64.75 to 37.48; P ⫽ 0.003). Average thinnest points were 443.6 m (range, 164 –587) for the Pentacam data and 445.9 m (range, 204 –597) for Orbscan IIz data. The mean difference in the thinnest point for the Pentacam and Orbscan IIz was ⫺2.28⫾35.55 (95% CI, ⫺14.31 to 9.75; 95% LOA, ⫺71.95 to 67.39; P ⫽ 0.703). The Bland–Altman plot for the mean difference is shown in Figure 3.
60 40 Highest elevation difference (Pentacam - Orbscan)
Radius difference (Pentacam - Orbscan)
0.80
37.48
20 0 -13.64
-20 -40 -60
-64.76
-80 -100 -120 -140 0
20
40
60
80
100
120
140
Highest elevation (average of Pentacam and Orbscan)
Figure 2. Bland–Altman plot of the mean difference of the elevation of the posterior curve generated by the Pentacam and Orbscan IIz. The mean difference of the highest elevation obtained was ⫺13.64⫾26.08 m (standard deviation) (95% confidence interval, ⫺22.46 to ⫺4.81; P ⫽ 0.003). The graph shows the 95% limits of agreement to be between ⫺64.76 and 37.48, indicating that there was a greater proportion of larger measurements by Orbscan IIz.
Quisling et al 䡠 Posterior Curvature Topography Measurements in Keratoconus Eyes 240
Thinnest point difference (Pentacam - Orbscan)
160
80
67.39
0
-2.28
-71.95
-80
-160
-240 160
240
320
400
480
560
640
Thinnest point (average of Pentacam and Orbscan)
Figure 3. Bland–Altman plot showing the difference between the thinnest point found by the Pentacam and Orbscan IIz. The mean difference in the thinnest point was ⫺2.28⫾35.55 m (95% confidence interval, ⫺14.31 to 9.75; P ⫽ 0.703). This graph shows that most of the difference between the measurements is no greater than ⫺71.95 and 67.39 (95% limits of agreement).
Discussion Corneal topography often aids in the diagnosis of keratoconus. The characteristic stromal thinning in keratoconus corresponds to an increased posterior elevation above the best-fit sphere on corneal topography. The Pentacam Comprehensive Eye Scanner is based on Scheimpflug technology, and the Orbscan is based on a scanning-slit beam with a Placido disk. Because Orbscan derives the posterior elevation map mathematically, there had been caution raised regarding the posterior curve in the past literature. This retrospective study found a statistically significant difference between posterior elevations above the best-fit sphere that Orbscan IIz and the Pentacam generated, despite no difference in the radii of curvature of the sphere. This difference may be related to the way the 2 machines obtain or interpret their data. Orbscan IIz uses slit scanning/Placido ring technology to measure the cornea surface. The first Orbscan topographer was based on a slit-scanning beam projected on the cornea.1 Then, in 1999 Orbscan IIz added a Placido image to obtain curvature measurements directly. Data points from the reflection of the mires and the 20 slit scans are used to make a representation of the anterior surface of the cornea. Orbscan IIz digitally recreates the posterior cornea, anterior iris, and lens from the collected data as previously described.1 This recreation requires sophisticated triangulation of the previously generated elevation and curvature of the anterior topography.1,3,4 Unlike Orbscan IIz, the Pentacam is based on a capture method called Scheimpflug imagery. By measuring 25 000 elevation points, it gives a representation of the shape of the
corneal surface. These elevation points are independent of axis, orientation, and position and yield only one possible curvature map. The difference in the calculated radii for the posterior best-fit sphere of the Pentacam and Orbscan IIz was not found to be statistically significant, which makes the difference in the measured posterior elevation between the machines important, as this could point to a difference in data analysis by the machines. Analogous to the anterior curve, Orbscan IIz may be more accurate in the periphery but less so in the center, thereby overestimating the posterior height. In the study, Orbscan IIz seems to generate an elevation above the best sphere that is greater than that of the Pentacam. Because most of the literature is based on Orbscan IIz, interpretation of the Pentacam should be based on individual experience with the machine. In our study, the pachymetry of the thinnest point calculated by the Pentacam and Orbscan IIz showed no significant difference in mean deviation. Orbscan has been shown to yield values for pachymetry similar to those of ultrasound in normal corneas and in keratoconus eyes.5,6 Both machines similarly found the location of the cones to be primarily in the inferior temporal quadrant, as in previous studies.7–10 There can be various reasons why there was a difference in the posterior elevation. Tear film abnormalities and the presence of anterior stromal scarring may contribute to this difference. In the small selection of patients who had both the scans present and diagnosis of keratoconus, some of the most disparate comparisons were in patients who had documented contact lens use. More studies will have to be performed in the future to elucidate this more clearly— possibly with a standardized test object. Both the Pentacam and Orbscan IIz provide useful information with regard to location of the keratoconus cone and pachymetry, despite obtaining the topography by different mechanisms. From these data, it appears that Orbscan determines the posterior elevation above the best-fit sphere higher than the Pentacam. This difference is probably due to the difference in methods between the 2 systems, as the scanning-slit topographer estimates the central 3 m, whereas the Scheimpflug images the cornea directly. But there are other mechanisms that could explain the disparity, including density of the pixels, the software used to image the cornea, and clarity of the cornea or tear film. The overall reason for the difference has yet to be elucidated, and more studies are needed. The importance of the study is to understand that when using the machines a direct comparison may not be possible, but the surgeon and clinician should become adept at interpreting the machine they use routinely.
References 1. Cairns G, McGhee CN. Orbscan computerized topography: attributes, applications, and limitations. J Cataract Refract Surg 2005;31:205–20. 2. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135– 60.
1631
Ophthalmology Volume 113, Number 9, September 2006 3. Snook RK. Pachymetry and true topography using the Orbscan system. In: Gills JP, Sanders DR, Thornton SP, Martin RG, eds. Corneal Topography: The State of the Art. Thorofare, NJ: SLACK Inc.; 1995:89 –103. 4. Cairns G, Collins A, McGhee CN. A corneal model for slit-scanning elevation topography. Ophthalmic Physiol Opt 2003;23;193–204. 5. Modis L Jr, Langenbucher A, Seitz B. Evaluation of normal corneas using the scanning-slit topography/pachymetry system. Cornea 2004;23:689 –94. 6. Gherghel D, Hosking SL, Mantry S, et al. Corneal pachymetry
1632
7. 8. 9. 10.
in normal and keratoconic eyes: Orbscan II versus ultrasound. J Cataract Refract Surg 2004;30:1272–7. Auffarth GU, Wang L, Volcker HE. Keratoconus evaluation using the Orbscan topography system. J Cataract Refract Surg 2000;26:222– 8. Maquire LJ, Bourne WM. Corneal topography of early keratoconus. Am J Ophthalmol 1989;108:107–12. Wilson SE, Klyce SD. Advances in the analysis of corneal topography. Surv Ophthalmol 1991;35:269 –77. Wilson SE, Lin DT, Klyce SD. Corneal topography of keratoconus. Cornea 1991;10:2– 8.