Accepted Manuscript Pentacam Scheimpflug tomography findings in topographically-normal patients and subclinical keratoconus cases Pablo R. Ruiseñor Vázquez, Jonatán David Galletti, Natalia Mínguez, Marianella Delrivo, Fernando Fuentes Bonthoux, Tomás Pförtner, Jeremías Gastón Galletti PII:
S0002-9394(14)00172-X
DOI:
10.1016/j.ajo.2014.03.018
Reference:
AJOPHT 8877
To appear in:
American Journal of Ophthalmology
Received Date: 4 February 2014 Revised Date:
27 March 2014
Accepted Date: 28 March 2014
Please cite this article as: Ruiseñor Vázquez PR, Galletti JD, Mínguez N, Delrivo M, Fuentes Bonthoux F, Pförtner T, Galletti JG, Pentacam Scheimpflug tomography findings in topographically-normal patients and subclinical keratoconus cases, American Journal of Ophthalmology (2014), doi: 10.1016/ j.ajo.2014.03.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Pentacam Scheimpflug tomography findings in topographically-normal patients and subclinical keratoconus cases
RI PT
Authors: Ruiseñor Vázquez Pablo R1,2, Galletti Jonatán David1,2, Mínguez Natalia1,2, Delrivo Marianella1, Fuentes Bonthoux Fernando1, Pförtner Tomás1, Galletti Jeremías Gastón1, 3
M AN U
Corresponding author: Jeremías Gastón Galletti
SC
1 ECOS (Clinical Ocular Studies) Laboratory, Buenos Aires, Argentina 2 Ophthalmology Division, Hospital de Clínicas José de San Martín, University of Buenos Aires, Argentina 3 Institute of Experimental Medicine, National Academy of Medicine/CONICET, Buenos Aires, Argentina
TE D
mailing address: ECOS (Clinical Ocular Studies) Laboratory (1119) Pueyrredón 1716 7 B Buenos Aires, Argentina phone/fax number: +54 (11) 48231189 email: [email protected]
AC C
EP
Short title: Pentacam keratoconus screening in refractive surgery
ACCEPTED MANUSCRIPT
Abstract Purpose: To evaluate Pentacam ectasia detection indices in topographically-normal patients and in subclinical keratoconus cases. Design: Prospective, observational case series.
RI PT
Methods: Setting: institutional; Patients: group 1 comprised one eye from 189 patients with unremarkable topography and groups 2 and 3 included the better and worse eyes, respectively, of 55 keratoconic patients. Group 2 eyes with normal topography (n=37) were considered subclinical keratoconus cases; Observation procedure: Pentacam Scheimpflug tomography; Main outcome measurements: 11 Pentacam ectasia detection indices.
M AN U
SC
Results: All Pentacam ectasia indices significantly differed between groups 1 and 2 and were correlated with keratoconus grade. Only 99 (52%) eyes in group 1 had normal values for every index, whereas 7 (19%) subclinical keratoconus eyes showed 2 or less abnormal indices. Standardized relational thickness and overall deviation indices had 73% and 89% sensitivity for subclinical keratoconus, respectively. Both average and maximum pachymetric progression indices offered 84% sensitivity while maximum relational thickness index showed 78% sensitivity for subclinical keratoconus. Optimized cutoff values for subclinical keratoconus increased the sensitivity of the standardized and maximum relational thickness indices.
AC C
EP
TE D
Conclusion: Pentacam Scheimpflug tomography can detect most subclinical keratoconus cases with unremarkable topography but performance is not as good as reported and varies considerably for each index. The overall deviation, average and maximum pachymetric progression and maximum relational thickness indices offer the highest sensitivity, which can be improved by using optimized cutoff values. Specificity constitutes an issue for some indices and up to 10% of subclinical keratoconus cases may go undetected by this technology.
ACCEPTED MANUSCRIPT
Introduction
RI PT
Moderate and severe keratoconus cases are relatively easy to identify because of the typical clinical signs and topographical features, but its subclinical forms (topographically-unremarkable corneas that later develop ectatic changes) represent a diagnostic challenge. Early keratoconus detection, especially in its subclinical form, increases the safety of corneal refractive surgery1 and allows for prompt treatment and prevention of ectatic disease. Placido disk-based corneal topography is commonly used for keratoconus diagnosis, but it is of little help with subclinical cases because of the lack of unequivocal findings when considered alone2. Fortunately there are several complementary techniques that can aid the physician in the detection of this disorder3–5, especially in its initial stages and in dubious cases.
AC C
EP
TE D
M AN U
SC
Scheimpflug tomography has become a widely employed technique for anterior eye imaging, probably because it conveys a wealth of information concerning corneal thickness and back surface3,6,7. As keratoconus-associated changes allegedly first arise in the posterior corneal surface8, it has been proposed that corneal Scheimpflug tomography could readily detect topographically-normal keratoconus cases6,7,9–17. To this end, the Pentacam system (a widely spread Scheimpflug tomograph) provides standardized indices for ectasia detection11 (at least 8 indices in software version 1.19r11), but their diagnostic performance has not been thoroughly explored when dealing with truly subclinical keratoconus eyes. More importantly, these indices yield information on complementary aspects of corneal shape but their comparative effectiveness has not been addressed. There is a real need for studies that mimic the clinical setting, where a refractive surgery candidate presents with an abnormal value in only one or two of the many keratoconus detection indices and the ophthalmologist must decide without knowing the relative importance of each corneal descriptor. Some of the published studies did not analyze a strictly subclinical sample of keratoconus cases12,13,15 or focused on particular aspects of the Pentacam’s Belin/Ambrosio Enhanced Ectasia display6,10,11,14,16. To address these issues, we set out in this study, on the one hand, to analyze Scheimpflug tomography findings in a large sample of healthy young patients with unremarkable corneal topography, and on the other hand, to assess the actual diagnostic yield of each of the Pentacam’s ectasia detection indices for truly subclinical keratoconus cases.
ACCEPTED MANUSCRIPT
Materials and Methods
M AN U
SC
RI PT
The study was an observational case series. The research protocol followed the tenets of the Declaration of Helsinki and was approved by the Hospital de Clínicas José de San Martín ethics committee. All subjects were told of the purpose of the study and gave written informed consent before inclusion. Patients were recruited between March and November 2013 at ECOS (Clinical Ocular Studies) Laboratory and had been referred for spectacle or contact lens prescription, preoperative screening or keratoconus diagnosis. Each subject underwent slit-lamp examination, Placido disk topography and aberrometry (iTrace, Tracey Technologies, Houston, TX, USA – software version 4.2.1) and Pentacam HR Scheimpflug tomography (Oculus Optikgerate GmbH, Wetzlar, Germany – software version 1.19r11). All measurements were performed by experienced operators (PRV, JDG and MD) in a darkened room within a 15-minute period, and the subjects were told to blink immediately before each examination. Only good-quality exams were accepted, defined as automated captures for the iTrace device in which all the reflected Placido rings were free of artifacts, and as automatic Scheimpflug scans (25 images in 2 seconds) that passed the Pentacam software’s quality check. For topography and keratoconus grading, the Keratoconus Severity Score was used18, which is based on average corneal power and corneal higher-order aberrations (expressed in µm as root-mean-square values). The Keratoconus Severity Score scale ranges from 0 (unaffected, normal topography), 1 (unaffected, atypical topography), 2 (suspect), 3 (mild keratoconus), 4 (moderate keratoconus) to 5 (severe keratoconus). Participant exclusion criteria were the following: previous eye surgery, any eye disease other than keratoconus, chronic use of topical medications or corneal opacities. Patients were asked to cease contact lens wear at least three weeks before examination.
AC C
EP
TE D
Three independent groups were recruited for this study. Patients with unremarkable slit lamp exam and normal topography (Keratoconus Severity Score 0: typical axial topography pattern, average corneal power ≤ 47.75 D and higher-order aberrations < 0.65 µm) in both eyes were recruited in group 1 (only one eye was randomly selected). Such strict topographical criteria were established to restrict the sample to patients without suspicious clinical and topographical findings and were adapted from Buhren et al19. Patients with biomicroscopy and topographical findings of manifest keratoconus (Keratoconus Severity Score 3 or greater: axial topography pattern consistent with keratoconus, may have positive slit-lamp findings, no corneal scarring, average corneal power > 49.00 D and higher-order aberrations > 1.50 µm) in at least one eye were considered keratoconic, and corneal curvature was used as a surrogate indicator of ectasia grade20: the eye with the lowest average corneal power was included in group 2 (better or less advanced keratoconic eyes) while the eye with the highest average corneal power was included in group 3 (worse or more advanced keratoconic eyes). No patient had exactly the same average corneal power in both eyes. Eyes in group 2 with insufficient topographical findings to meet diagnostic criteria for keratoconus (Keratoconus Severity Score 0-2) were considered subclinical cases and analyzed further as subgroup 2SC. The following corneal descriptors were obtained from the Pentacam’s software Belin-Ambrósio Enhanced Ectasia Display: central and thinnest pachymetry, front and back corneal elevation at the thinnest corneal point, pachymetric progression indices (minimum, average and maximum), normalized indices Df (deviation of normality of the front elevation), Db (deviation of normality of the back elevation), Dp (deviation of normality of pachymetric progression), Dt (deviation of normality of corneal thinnest point), Da (deviation of normality of relational thickness) and D
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
(overall deviation of normality), and the Ambrosio’s average and maximum relational thickness indices. The methodology for calculation has been disclosed by the manufacturer for some but not all of these indices6,11. Pachymetric progression refers to the percentage in corneal thickness increase along each meridian starting from the thinnest corneal point, and the minimum, average and maximum values are reported, whereas relational thickness indices (average and maximum) express the ratio of the thinnest pachymetry and the respective pachymetric progression. Upon request, the manufacturer informed that the deviation indices, average pachymetric progression and maximum relational thickness are classified by the software as normal (< 1.6 standard deviation –SD- from the population mean, shown in white), suspicious (≥ 1.6 SD and < 2.6 SD, highlighted in yellow) and pathologic (≥ 2.6 SD, highlighted in red), from the data reported by Ambrosio et al11, and this scheme was followed throughout this work. For the normalized indices, each observation was classified into one of the three categories, whereas for the pachymetric progression and relational thickness indices, each observation was categorized as normal or abnormal according to the cutoff values reported by Ambrosio et al11: minimum pachymetric progression ≥ 0.79, average pachymetric progression ≥ 1.06, maximum pachymetric progression ≥ 1.44, average relational thickness ≤ 424 and maximum relational thickness ≤ 339 were considered abnormal.
AC C
EP
TE D
Normality of the data was assessed by the Kolmogorov–Smirnov test, and parametric or nonparametric tests were subsequently used. Receiver operating characteristic (ROC) curves were used to calculate sensitivity, specificity and area under the curve (AROC) of each corneal index, as previously described4. Optimal cutoff points for each index were obtained from the ROC curves as those closest to the perfect classification point4. Logistic regression with forward stepwise inclusion (logistic function 1: logit = 0.467 x Db + 2.625 x Da – 4.871) and all variables entered at once (logistic function 2: logit = 0.333 x Df + 0.353 x Db – 0.466 x Dp + 0.077 x Dt + 3.149 x Da – 5.257) was employed to combine the 5 individual D indices, with a 0.5 cutoff. Statistical tests and analysis were performed with Prism 5 software (GraphPad Software, La Jolla, CA, USA) and SPSS 17 software (SPSS Inc., Chicago, IL, USA). Statistical significance was set at p < 0.05 and data are shown as mean±standard deviation unless otherwise stated. Data collection and sorting were done with the aid of Microsoft Excel 2010 software.
ACCEPTED MANUSCRIPT
Results Demographics and keratoconus grade distribution
Corneal descriptors in control and keratoconic eyes
SC
RI PT
Group 1 (control eyes) comprised 189 eyes (94 right and 95 left) from 189 patients with unremarkable topography (Keratoconus Severity Score 0) in both eyes. The less advanced eye (29 right and 26 left) of 55 keratoconus patients was included in group 2 whereas the worse eye (26 right and 29 left) was included in group 3. There was no significant difference in mean age between control (32.3±8.1 years, range 14-71) and keratoconic (32.5±11.7 years, range 14-65) patients, but gender distribution significantly differed (41% vs 60% males, p=0.01) between groups. According to Keratoconus Severity Score grading, group 2 included 21 (38%) grade 0 eyes, 7 (13%) grade 1 eyes, 9 (16%) grade 2 eyes, 15 (27%) grade 3 eyes, 1 (1%) grade 4 eye and 2 (4%) grade 5 eyes, whereas group 3 comprised 33 (60%) grade 3 eyes, 11 (20%) grade 4 eyes and 11 (20%) grade 5 eyes.
M AN U
Conventional corneal descriptors are summarized in Table 1 and depicted in Figure 1. Means for every Pentacam corneal descriptor in groups 2 and 3 were significantly different when compared with group 1 (Table 1, Figures 2 and 3). In group 2 eyes, every keratoconus index was significantly correlated with Keratoconus Severity Score grading (Spearman’s correlation, Table 1 for r values, p < 0.001). Truly subclinical keratoconus eyes (n=37 [67%], group 2 eyes with Keratoconus Severity Score 0-2) were analyzed further as subgroup 2SC (Table 1), and they also differed significantly in means with group 1 eyes for every Pentacam corneal descriptor analyzed.
TE D
Classification of control eyes according to their Pentacam corneal indices
AC C
EP
The classification results according to each keratoconus detection index are summarized in Table 2 and Figure 4. In group 1, 129 eyes (68.3%) had normal values for every D index, and of these, all but three (126 eyes, 66.7%) showed also normal values for maximum relational thickness and all but 21 (108 eyes, 57.1%) also exhibited normal average pachymetric progression values. When considering every deviation index and maximum relational thickness and average pachymetric progression altogether, 107 eyes (56.6%) showed normal values for every index. Regarding the pachymetric progression indices, only 128 eyes (67.7%) exhibited normal values in all of them, whereas for average and maximum relational thickness, 169 eyes (89.4%) had normal values in both indices. Only 99 eyes (52.4%) in group 1 did not show abnormal scores in each of the 11 indices reported by the Pentacam. Group 1 eyes with normal overall deviation values (n=152, 80.4%) did not significantly differ in higher-order aberrations with those with suspicious or abnormal values (0.250±0.072 vs 0.256±0.060 µm, p=0.66). Group 1 eyes with normal pachymetric progression deviation values (n=173, 91.5%) also did not differ in higherorder aberrations when compared with those group 1 eyes flagged as suspicious or abnormal (≥1.6 SD) according to this metric. Keratoconus detection by Pentacam corneal indices In group 3 (worse eyes), all observations had 7 or more abnormal values in the 11 keratoconus detection indices evaluated, whereas in group 2 (better eyes), 2 (3.6%) observations exhibited no abnormal indices and 7 (12.7%) observations had 2 or less abnormal indices (Figure 4). Considering a suspicious value (≥1.6 SD) as positive (Table 2 and Figure 2), the sensitivity of
ACCEPTED MANUSCRIPT
RI PT
front, back, pachymetric progression, corneal thickness, relational thickness and overall deviation indices for group 2 eyes was 70.9%, 69.1%, 78.2%, 58.2%, 81.8% and 92.7%, respectively, whereas for subclinical keratoconus eyes (subgroup 2SC) the sensitivity was 56.8%, 54.1%, 67.6%, 43.2%, 73.0% and 89.2%, respectively. Within group 2, true positive cases for every deviation index exhibited greater mean higher-order aberrations than false negative cases, and compared with group 1 eyes, mean higher-order aberrations were significantly increased in false negative eyes for all metrics except for the overall deviation index. Considering the previously reported cutoff values (Table 2 and Figure 3), the sensitivity of the minimum, average and maximum pachymetric progression and average and maximum relational thickness indices for group 2 eyes was 76.4%, 89.1%, 89.1%, 80.0% and 85.5%, respectively, whereas for subclinical keratoconus eyes (subgroup 2SC) was 64.9%, 83.8%, 83.8%, 70.3% and 78.4%, respectively.
AC C
EP
TE D
M AN U
SC
A logistic regression approach with the 5 individual deviation indices selected the combination of back and relational thickness indices as the best possible. This diagnostic function predicted 186 eyes in group 1 (98.4%) as healthy and 44 eyes in group 2 (80.0%) as keratoconic, whereas a logistic regression approach with the 5 individual deviation indices entered at once predicted 187 eyes in group 1 (98.9%) as healthy and 44 eyes in group 2 (80.0%) as keratoconic. In comparison, the Pentacam’s overall deviation index with a ≥2.6 SD cutoff marked 186 eyes in group 1 as healthy and 44 eyes in group 2 as keratoconic, and the classification matched that of the first logistic function except for one case in the keratoconus group. An independent ROC analysis was performed between groups 1 and 2 in order to validate the previously published cutoff values for each Pentacam corneal index, and the sensitivity and specificity are summarized in Table 3 along with optimal cutoff values derived from these samples.
ACCEPTED MANUSCRIPT
Discussion
RI PT
Pentacam Scheimpflug tomography provides specific corneal analysis indices designed to diagnose ectatic changes. In this work we explored the actual performance of these descriptors in a large sample of topographically-normal patients and in keratoconus cases to assess which indices are more adequate for diagnosing this disorder at its initial stage. The results altogether show that Scheimpflug tomography allows the clinician to find early ectatic changes, and at the same time, that there is still a considerable fraction of subclinical keratoconus cases that can go undetected even by this method.
M AN U
SC
In agreement with previous reports that utilized diverse criteria for defining keratoconus6,7,9–16, we observed that all Pentacam indices differed significantly between healthy eyes and the better eyes of keratoconus patients diagnosed by the findings in their more advanced fellow eyes. In this keratoconic sample, two out of three eyes (n=37, 67%) could be considered as subclinical cases, that is, with insufficient topography findings to meet the Keratoconus Severity Score criteria for manifest disease, and most of this subset (21 eyes, 57%) had unremarkable Placido topography. Even this stricter subset had reduced corneal thickness, increased anterior surface irregularity and showed significant differences in every Pentacam index, suggesting that at least slight changes in corneal shape were present in these keratoconic eyes with subclinical disease, as others have shown2,19. There was however considerable overlap in Pentacam indices between group 1 eyes and keratoconic eyes (Figures 2 and 3), which affected their diagnostic performance.
AC C
EP
TE D
Regarding the standardized indices, which are the ones that offer the most straightforward interpretation in the Pentacam report, the diagnostic yield varied markedly between each metric. Although increased posterior elevation has been proposed as a sensitive marker for keratoconus10,21, back corneal surface deviation index had poor performance for subclinical keratoconus detection and was not clinically superior to anterior corneal surface deviation, as both indices flagged only about one half of the subclinical keratoconus eyes as abnormal (Table 2). Our findings suggest that detectable back corneal surface abnormalities do not necessarily arise before anterior changes. This result, however, differs with that reported by de Sanctis et al10,21, which was still low (68-73% sensitivity), and probably derives from the more stringent diagnostic criteria that we employed for defining subclinical disease. Supporting this idea, the sensitivity of posterior corneal elevation in our sample was about 67% when we considered the better fellow eyes of manifest keratoconus cases (group 2). Corneal thickness deviation had the lowest sensitivity (43%) among the studied indices, and this result simply reflects the significant overlap in pachymetric values between healthy eyes and truly subclinical keratoconus cases14,16. In contrast to front and back elevation, focalized corneal thinning seems to typify some of the earliest changes in keratoconic eyes. The Pentacam software provides several metrics designed to detect these ectatic signs: pachymetric progression and Ambrosio’s relational thickness indices6,11, along with standardized versions to ease interpretation. According to the reported cutoff points11, average and maximum pachymetric progression were the single metrics with the highest sensitivity (83.8%) for subclinical keratoconus, and maximum relational thickness followed closely in our sample. Nevertheless their performance was not as good as initially reported and probably reflects the fact that we tested them on a subclinical keratoconus sample instead of patients with bilateral disease11. Remarkably, the respective standardized indices exhibited lower sensitivity than their unstandardized counterparts, and this was especially true for standardized pachymetric progression, which only detected about two thirds of the subclinical
ACCEPTED MANUSCRIPT
RI PT
keratoconus eyes in our sample. An independent ROC analysis (Table 3) yielded optimal cutoff values that were similar to those initially reported, albeit slightly displaced towards the normal mean for every metric, evidencing the different composition of the keratoconic sample. The optimized cutoff value of 1.26) conveyed comparable levels of detection, but this was not true for the standardized pachymetric progression index. It should be noted that these findings still require independent validation as they might reflect the specific composition of our sample.
EP
TE D
M AN U
SC
As no specific cutoff value for each individual deviation index completely separated both groups, a multivariate approach that combined several indices might yield better diagnostic performance. The Pentacam software provides such metric, the overall deviation index, although the exact statistical method underlying its calculation was not disclosed by the manufacturer and its actual performance has not been independently validated. In our sample this index correctly identified 89.2% of the subclinical keratoconus cases, providing one of the highest sensitivities among the Pentacam’s metrics. We conducted our own analysis of combinations of the deviation indices, and either by selecting the best possibility with the fewest indices (back and relational thickness deviation) or the best possibility with every index we obtained almost identical results to those of the overall deviation index with a ≥2.6 SD cutoff, which is the actual value reported as pathologic by the Pentacam software. We infer from this analysis that the unknown algorithm built in the software already provides optimal results, and more importantly, that there is a limit to the diagnostic capacity of the slight changes in corneal shape detected by Scheimpflug tomography. We chose to consider a suspicious value (≥1.6 SD) as positive in order to maximize sensitivity while sacrificing specificity; in other words, because we believe that in the preoperative evaluation of refractive surgery candidates it is preferable to falsely flag a cornea as ectatic than to miss a subclinical keratoconus case. The Pentacam’s suspicious cutoff value for the overall deviation index was validated by our independent ROC analysis, which suggested >1.61 as optimal for this particular keratoconic sample. This approach, however, predicted about one every five eyes in the control group to be potentially keratoconic, an excessively high false positive rate (specificity 82.3%). In contrast, modified cutoff values for standardized and maximum relational thickness offered comparable sensitivity and improved specificity in our sample (Table 3), and perhaps clinicians should rely more on these indices to reduce the false positive rate. These findings nevertheless await independent validation.
AC C
There is one considerable limitation in this study, which is the composition of the control group. As we attempted to reproduce the clinical setting in which Pentacam analysis takes place, only clinical and topographical data gathered at the same visit were considered. In this way we followed similar criteria to other reports, which include standard Keratoconus Severity Score criteria for unremarkable slit lamp exam and corneal topography in both eyes. Nonetheless a significant proportion of subclinical keratoconus cases have normal topography (defined as Keratoconus Severity Score 0), as evidenced by the group 2 keratoconic eyes, and considering the large sample size of the control group (n=189) and the keratoconus prevalence reported elsewhere13, it is expected that a few subclinical keratoconus cases would be artificially included in this group and therefore specificity data from it derived could be inaccurate. This seems to be the case for at least three observations that showed markedly abnormal overall deviation values. The remaining 34 cases with a suspicious overall deviation value probably comprise false
ACCEPTED MANUSCRIPT
positives that will never develop ectatic changes and some early keratoconus cases, but prospective studies are required to properly address this matter.
AC C
EP
TE D
M AN U
SC
RI PT
In conclusion, we found that Pentacam Scheimpflug tomography analysis can detect most but not all subclinical keratoconus cases with insufficient anterior surface findings to be considered as keratoconic according to the Keratoconus Severity Score scale. As provided by the software, the multivariate overall deviation metric was most appropriate for this task, whereas the performance of the individual indices fell considerably behind unless modified cutoff values were used. We also observed that when evaluating apparently healthy subjects with unremarkable topography, there are a sizable number of cases that are flagged as suspects by the Pentacam analysis, and that additional testing should perhaps be applied to these patients. There are other keratoconus detection tools, such as corneal biomechanical testing4,22 and epithelial thickness analysis5,23, which rely on different concepts and could be used complementarily with Scheimpflug tomography.
ACCEPTED MANUSCRIPT
Acknowledgements
AC C
EP
TE D
M AN U
SC
RI PT
a) Funding/Support: None b) Financial Disclosures: None c) Author Contributions: design (FFB, TP, JGG) and conduct of the study (PRV, JDG, NM and MD); collection (PRV, JDG, NM and MD), analysis (JGG), and interpretation of the data (PRV, JDG and JGG); and preparation (PRV, JDG and JGG), review (NM), and approval of the manuscript (FFB and TP). d) Other Acknowledgments: None
ACCEPTED MANUSCRIPT
References
6.
7.
8.
9.
10.
11. 12. 13.
14.
15.
16.
RI PT
SC
5.
M AN U
4.
TE D
3.
EP
2.
Klein SR, Epstein RJ, Randleman JB, Stulting RD. Corneal ectasia after laser in situ keratomileusis in patients without apparent preoperative risk factors. Cornea. 2006;25(4):388–403. Bühren J, Schäffeler T, Kohnen T. Preoperative topographic characteristics of eyes that developed postoperative LASIK keratectasia. J Refract Surg. 2013;29(8):540–9. Belin MW, Ambrósio R. Scheimpflug imaging for keratoconus and ectatic disease. Indian J Ophthalmol. 2013;61(8):401–6. Galletti JG, Pförtner T, Fuentes Bonthoux F. Improved keratoconus detection by ocular response analyzer testing after consideration of corneal thickness as a confounding factor. J Refract Surg. 2012;28(3):202–8. Reinstein DZ, Archer TJ, Gobbe M. Stability of LASIK in topographically suspect keratoconus confirmed non-keratoconic by Artemis VHF digital ultrasound epithelial thickness mapping: 1-year follow-up. J Refract Surg. 2009;25(7):569–77. Ambrósio R, Alonso RS, Luz A, Coca Velarde LG. Corneal-thickness spatial profile and corneal-volume distribution: tomographic indices to detect keratoconus. J Cataract Refract Surg. 2006;32(11):1851–9. Smadja D, Santhiago MR, Mello GR, Krueger RR, Colin J, Touboul D. Influence of the reference surface shape for discriminating between normal corneas, subclinical keratoconus, and keratoconus. J Refract Surg. 2013;29(4):274–81. Piñero DP, Alió JL, Alesón A, Escaf Vergara M, Miranda M. Corneal volume, pachymetry, and correlation of anterior and posterior corneal shape in subclinical and different stages of clinical keratoconus. J Cataract Refract Surg. 2010;36(5):814–25. Nilforoushan M-R, Speaker M, Marmor M, et al. Comparative evaluation of refractive surgery candidates with Placido topography, Orbscan II, Pentacam, and wavefront analysis. J Cataract Refract Surg. 2008;34(4):623–31. De Sanctis U, Loiacono C, Richiardi L, Turco D, Mutani B, Grignolo FM. Sensitivity and specificity of posterior corneal elevation measured by Pentacam in discriminating keratoconus/subclinical keratoconus. Ophthalmology. 2008;115(9):1534–9. Ambrósio R, Caiado ALC, Guerra FP, et al. Novel pachymetric parameters based on corneal tomography for diagnosing keratoconus. J Refract Surg. 2011;27(10):753–8. Miháltz K, Kovács I, Takács A, Nagy ZZ. Evaluation of keratometric, pachymetric, and elevation parameters of keratoconic corneas with pentacam. Cornea. 2009;28(9):976–80. Serdarogullari H, Tetikoglu M, Karahan H, Altin F, Elcioglu M. Prevalence of keratoconus and subclinical keratoconus in subjects with astigmatism using pentacam derived parameters. J Ophthalmic Vis Res. 2013;8(3):213–9. Uçakhan ÖÖ, Cetinkor V, Özkan M, Kanpolat A. Evaluation of Scheimpflug imaging parameters in subclinical keratoconus, keratoconus, and normal eyes. J Cataract Refract Surg. 2011;37(6):1116–24. Ahmadi Hosseini SM, Abolbashari F, Niyazmand H, Sedaghat MR. Efficacy of corneal tomography parameters and biomechanical characteristic in keratoconus detection. Cont Lens Anterior Eye. 2014;37(1):26–30. Muftuoglu O, Ayar O, Ozulken K, Ozyol E, Akıncı A. Posterior corneal elevation and back difference corneal elevation in diagnosing forme fruste keratoconus in the fellow eyes of unilateral keratoconus patients. J Cataract Refract Surg. 2013;39(9):1348–57.
AC C
1.
ACCEPTED MANUSCRIPT
23.
RI PT
22.
SC
21.
M AN U
20.
TE D
19.
EP
18.
Ambrósio R, Nogueira LP, Caldas DL, et al. Evaluation of corneal shape and biomechanics before LASIK. Int Ophthalmol Clin. 2011;51(2):11–38. McMahon TT, Szczotka-Flynn L, Barr JT, et al. A new method for grading the severity of keratoconus: the Keratoconus Severity Score (KSS). Cornea. 2006;25(7):794–800. Buhren J, Kuhne C, Kohnen T. Defining subclinical keratoconus using corneal firstsurface higher-order aberrations. Am J Ophthalmol. 2007;143(3):381–9. Zadnik K, Barr JT, Edrington TB, et al. Baseline findings in the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) Study. Invest Ophthalmol Vis Sci. 1998;39(13):2537–46. De Sanctis U, Aragno V, Dalmasso P, Brusasco L, Grignolo F. Diagnosis of subclinical keratoconus using posterior elevation measured with 2 different methods. Cornea. 2013;32(7):911–5. Ruiseñor Vázquez PR, Delrivo M, Fuentes Bonthoux F, Pförtner T, Galletti JG. Combining Ocular Response Analyzer Metrics for Corneal Biomechanical Diagnosis. J Refract Surg. 2013;29(9):596-602. Rocha KM, Perez-Straziota CE, Perez-Straziota E, Stulting RD, Randleman JB. SD-OCT analysis of regional epithelial thickness profiles in keratoconus, postoperative corneal ectasia, and normal eyes. J Refract Surg. 2013;29(3):173–9.
AC C
17.
ACCEPTED MANUSCRIPT
Figure Captions
RI PT
Figure 1. Conventional corneal descriptors in control and keratoconus eyes. Box-whisker plots of average corneal power (ACP, expressed in diopters –D-), corneal higher-order aberrations (HOA), central corneal thickness (CCT), thinnest pachymetry (Thinnest pach.) and front and back elevation at the thinnest corneal location in control eyes (group 1, empty boxes) and better (group 2, light-gray boxes) and worse (group 3, dark-gray boxes) eyes of keratoconus patients. * indicates a statistically-significant difference in mean compared with the control group (analysis of variance with Dunnett’s post hoc test or non-parametric equivalent).
M AN U
SC
Figure 2. Standardized deviation indices in control and keratoconus eyes. Box-whisker plots of deviation of front (Df) and back (Db) elevation, pachymetric progression (Dp), corneal thickness (Dt), relational thickness (Da) and combined index (D) in control eyes (group 1, empty boxes) and better (group 2, light-gray boxes) and worse (group 3, dark-gray boxes) eyes of keratoconus patients. The reported suspect (≥1.6 standard deviation –SD=) and abnormal (≥2.6 SD) cutoff values are plotted for reference. * indicates a statistically-significant difference in mean compared with the control group (analysis of variance with Dunnett’s post hoc test or nonparametric equivalent).
TE D
Figure 3. Pachymetric progression and relational thickness indices in control and keratoconus eyes. Box-whisker plots of minimum (PPImin), average (PPIave) and maximum (PPImax) pachymetric progression indices and average (ARTave) and maximum (ARTmax) Ambrosio’s relational thickness indices in control eyes (group 1, empty boxes) and better (group 2, light-gray boxes) and worse (group 3, dark-gray boxes) eyes of keratoconus patients. The reported cutoff values are plotted for reference. * indicates a statistically-significant difference in mean compared with the control group (analysis of variance with Dunnett’s post hoc test or nonparametric equivalent).
AC C
EP
Figure 4. Distribution of Pentacam’s keratoconus detection indices in control and keratoconus eyes. Frequency distribution (as percentage of total cases) of the number of abnormal indices (6 deviation indices, 3 pachymetric progression indices and 2 relational thickness indices) in control eyes (group 1), subclinical keratoconus eyes (subgroup 2SC) and better (group 2) and worse (group 3) eyes of keratoconus patients.
ACCEPTED MANUSCRIPT
Pentacam’s keratoconus indices
PPImax Df
Db Dp Dt
Da D
-0.21±0.90 (-1.49-4.05) 0.44±0.92 (-1.99-3.63) 0.33±0.96 (-2.16-3.09)
2.00±2.31# (-1.30-7.58) 3.49±3.24# (-1.05-13.94) 1.64±1.31# (-0.91-4.84)
4.15±4.44# (-1.30-23.22) , r=0.808 5.97±5.09# (-1.05-18.15), r=0.772 2.31±1.85# (-0.91-9.28), r=0.479
0.39±0.79 (-2.62-2.29) 1.03±0.70 (-0.87-4.86) 555±98 (333-893) 445±87 (237-770)
2.08±0.94# (-1.20-3.74) 4.26±3.03# (0.55-16.55) 381±125# (160-744) 269±101# (97-620)
2.48±1.00# (-1.20-3.99), r=0.664 6.08±3.98# (0.55-16.55), r=0.757 318±142# (112-744), r=-0.757 223±109# (78-620), r=-0.729
AC C
ARTave (µm) ARTmax (µm)
Keratoconus patients (n=55) Better keratoconus eyes Group 2 (n=55)
Worse keratoconus eyes Group 3 (n=55) 51.53±7.48# (42.23-77.24) 3.261±5.918# (0.373-36.840) 479±76# (321-940) 443±45# (298-550), r=0.415 26±15# (-11-72), r=0.676 54±26# (-1-130), r=0.667 1.90±0.86# (0.59-4.28), r=0.695 2.53±0.99# (1.03-5.61), r=0.741 3.70±1.53# (1.41-8.35), r=0.670 11.68±8.25# (-1.86-35.29), r=0.821 9.63±6.98# (-1.03-29.83),r=0.854 11.01±6.62# (0.84-31.85), r=0.736 3.50±2.06# (-0.35-12.46), r=0.419 3.21±0.96# (1.00-8.28), r=0.636 9.81±4.70# (1.42-22.92), r=0.833 208±99# (53-518), r=-0.704 145±73# (42-379), r=-0.662
SC
RI PT
46.03±3.50# (40.76-56.56) 1.153±0.947# (0.183-5.596) 491±39# (407-587), r=0.364 474±43# (376-572), r=0.474 16±12# (-3-43), r=0.779 36±27# (0-90), r=0.757 1.34±0.70# (0.46-3.48), r=0.722 1.79±0.77# (0.75-3.67), r=0.773 2.61±1.17# (0.90-5.35), r=0.743 5.43±5.18# (-1.19-21.97) , r=0.850
M AN U
PPIave
Subclinical keratoconus eyes Group 2SC (n=37) 44.71±2.19 (40.76-48.90) 0.669±0.350# (0.183-1.472) 502±37# (412-587) 489±37# (409-572) 10±8# (-3-36) 24±21# (0-85) 1.02±0.41# (0.46-2.25) 1.42±0.48# (0.75-2.97) 2.06±0.81# (0.90-4.90), r=0.464 2.71±2.48# (-1.19-8.67)
TE D
ACP (D) HOA (µm) CCT (µm) TP (µm) Front elevation at thinnest location Back elevation at thinnest location PPImin
Control patients Group 1 (n=189) 44.48±1.48 (40.83-47.48) 0.251±0.070 (0.145-0.497) 532±33 (449-628) 528±32 (448-625) 3±2 (-3-14) 4±5 (-7-37) 0.70±0.13 (0.37-1.16) 0.97±0.14 (0.61-1.44) 1.22±0.20 (0.73-2.02) 0.44±1.07 (-1.79-4.13)
EP
Conventional corneal descriptors
Table 1 – Topography & Pentacam corneal descriptors in control and keratoconic eyes
Data is shown as mean±standard deviation (minimum-maximum). 2SC = group 2 eyes with Keratoconus Severity Score 0; ACP = average corneal power; HOA = higherorder aberrations; CCT = central corneal thickness; TP = thinnest pachymetry; PPI = pachymetric progression indices (min = minimum, ave = average, max = maximum); Df = deviation of front elevation; Db = deviation of back elevation; Dp = deviation of pachymetric progression; Dt = deviation of corneal thickness; Da = deviation of relational thickness; D = overall deviation; ART = Ambrosio’s relational thickness indices (ave = average, max = maximum). # Statistically-significant difference with group 1 eyes.
ACCEPTED MANUSCRIPT
Table 2 – Classification table for Pentacam keratoconus indices Control eyes Group 1 (n=189) Abnorma Normal Suspect l (1.29
67.6
76.9
>1.00
89.2
90.3
>1.26
89.2
82.3
>1.61
73.0
73.7
78.4
82.8
86.5
87.1
88.9
78.4
90.5
86.5
AROC 0.86 (0.80-0.93) 0.86 (0.80-0.93) 0.91 (0.85-0.96) 0.85 (0.78-0.91) 0.96 (0.92-0.99) 0.95 (0.92-0.99) 0.85 (0.78-0.93) 0.90 (0.85-0.96) 0.95 (0.91-0.99) 0.92 (0.87-0.97) 0.95 (0.91-0.99)
>0.76
>1.09
>1.41 0.77
85.5
83.3
>1.29
72.7
82.3
>1.21
SC
Db
0.80 (0.71-0.89) 0.80 (0.71-0.89) 0.86 (0.78-0.94) 0.79 (0.71-0.88) 0.94 (0.88-0.99) 0.93 (0.88-0.98) 0.79 (0.68-0.89) 0.86 (0.78-0.94) 0.92 (0.86-0.98) 0.88 (0.80-0.95) 0.93 (0.87-0.98)
Spec
92.7
90.3
>1.26
85.5
95.7
>2.17
74.5
89.8
>0.85
85.5
82.2
>1.09
89.1
90.3
>1.46
88.9
85.5
S0002-9394(14)00172-X
DOI:
10.1016/j.ajo.2014.03.018
Reference:
AJOPHT 8877
To appear in:
American Journal of Ophthalmology
Received Date: 4 February 2014 Revised Date:
27 March 2014
Accepted Date: 28 March 2014
Please cite this article as: Ruiseñor Vázquez PR, Galletti JD, Mínguez N, Delrivo M, Fuentes Bonthoux F, Pförtner T, Galletti JG, Pentacam Scheimpflug tomography findings in topographically-normal patients and subclinical keratoconus cases, American Journal of Ophthalmology (2014), doi: 10.1016/ j.ajo.2014.03.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Pentacam Scheimpflug tomography findings in topographically-normal patients and subclinical keratoconus cases
RI PT
Authors: Ruiseñor Vázquez Pablo R1,2, Galletti Jonatán David1,2, Mínguez Natalia1,2, Delrivo Marianella1, Fuentes Bonthoux Fernando1, Pförtner Tomás1, Galletti Jeremías Gastón1, 3
M AN U
Corresponding author: Jeremías Gastón Galletti
SC
1 ECOS (Clinical Ocular Studies) Laboratory, Buenos Aires, Argentina 2 Ophthalmology Division, Hospital de Clínicas José de San Martín, University of Buenos Aires, Argentina 3 Institute of Experimental Medicine, National Academy of Medicine/CONICET, Buenos Aires, Argentina
TE D
mailing address: ECOS (Clinical Ocular Studies) Laboratory (1119) Pueyrredón 1716 7 B Buenos Aires, Argentina phone/fax number: +54 (11) 48231189 email: [email protected]
AC C
EP
Short title: Pentacam keratoconus screening in refractive surgery
ACCEPTED MANUSCRIPT
Abstract Purpose: To evaluate Pentacam ectasia detection indices in topographically-normal patients and in subclinical keratoconus cases. Design: Prospective, observational case series.
RI PT
Methods: Setting: institutional; Patients: group 1 comprised one eye from 189 patients with unremarkable topography and groups 2 and 3 included the better and worse eyes, respectively, of 55 keratoconic patients. Group 2 eyes with normal topography (n=37) were considered subclinical keratoconus cases; Observation procedure: Pentacam Scheimpflug tomography; Main outcome measurements: 11 Pentacam ectasia detection indices.
M AN U
SC
Results: All Pentacam ectasia indices significantly differed between groups 1 and 2 and were correlated with keratoconus grade. Only 99 (52%) eyes in group 1 had normal values for every index, whereas 7 (19%) subclinical keratoconus eyes showed 2 or less abnormal indices. Standardized relational thickness and overall deviation indices had 73% and 89% sensitivity for subclinical keratoconus, respectively. Both average and maximum pachymetric progression indices offered 84% sensitivity while maximum relational thickness index showed 78% sensitivity for subclinical keratoconus. Optimized cutoff values for subclinical keratoconus increased the sensitivity of the standardized and maximum relational thickness indices.
AC C
EP
TE D
Conclusion: Pentacam Scheimpflug tomography can detect most subclinical keratoconus cases with unremarkable topography but performance is not as good as reported and varies considerably for each index. The overall deviation, average and maximum pachymetric progression and maximum relational thickness indices offer the highest sensitivity, which can be improved by using optimized cutoff values. Specificity constitutes an issue for some indices and up to 10% of subclinical keratoconus cases may go undetected by this technology.
ACCEPTED MANUSCRIPT
Introduction
RI PT
Moderate and severe keratoconus cases are relatively easy to identify because of the typical clinical signs and topographical features, but its subclinical forms (topographically-unremarkable corneas that later develop ectatic changes) represent a diagnostic challenge. Early keratoconus detection, especially in its subclinical form, increases the safety of corneal refractive surgery1 and allows for prompt treatment and prevention of ectatic disease. Placido disk-based corneal topography is commonly used for keratoconus diagnosis, but it is of little help with subclinical cases because of the lack of unequivocal findings when considered alone2. Fortunately there are several complementary techniques that can aid the physician in the detection of this disorder3–5, especially in its initial stages and in dubious cases.
AC C
EP
TE D
M AN U
SC
Scheimpflug tomography has become a widely employed technique for anterior eye imaging, probably because it conveys a wealth of information concerning corneal thickness and back surface3,6,7. As keratoconus-associated changes allegedly first arise in the posterior corneal surface8, it has been proposed that corneal Scheimpflug tomography could readily detect topographically-normal keratoconus cases6,7,9–17. To this end, the Pentacam system (a widely spread Scheimpflug tomograph) provides standardized indices for ectasia detection11 (at least 8 indices in software version 1.19r11), but their diagnostic performance has not been thoroughly explored when dealing with truly subclinical keratoconus eyes. More importantly, these indices yield information on complementary aspects of corneal shape but their comparative effectiveness has not been addressed. There is a real need for studies that mimic the clinical setting, where a refractive surgery candidate presents with an abnormal value in only one or two of the many keratoconus detection indices and the ophthalmologist must decide without knowing the relative importance of each corneal descriptor. Some of the published studies did not analyze a strictly subclinical sample of keratoconus cases12,13,15 or focused on particular aspects of the Pentacam’s Belin/Ambrosio Enhanced Ectasia display6,10,11,14,16. To address these issues, we set out in this study, on the one hand, to analyze Scheimpflug tomography findings in a large sample of healthy young patients with unremarkable corneal topography, and on the other hand, to assess the actual diagnostic yield of each of the Pentacam’s ectasia detection indices for truly subclinical keratoconus cases.
ACCEPTED MANUSCRIPT
Materials and Methods
M AN U
SC
RI PT
The study was an observational case series. The research protocol followed the tenets of the Declaration of Helsinki and was approved by the Hospital de Clínicas José de San Martín ethics committee. All subjects were told of the purpose of the study and gave written informed consent before inclusion. Patients were recruited between March and November 2013 at ECOS (Clinical Ocular Studies) Laboratory and had been referred for spectacle or contact lens prescription, preoperative screening or keratoconus diagnosis. Each subject underwent slit-lamp examination, Placido disk topography and aberrometry (iTrace, Tracey Technologies, Houston, TX, USA – software version 4.2.1) and Pentacam HR Scheimpflug tomography (Oculus Optikgerate GmbH, Wetzlar, Germany – software version 1.19r11). All measurements were performed by experienced operators (PRV, JDG and MD) in a darkened room within a 15-minute period, and the subjects were told to blink immediately before each examination. Only good-quality exams were accepted, defined as automated captures for the iTrace device in which all the reflected Placido rings were free of artifacts, and as automatic Scheimpflug scans (25 images in 2 seconds) that passed the Pentacam software’s quality check. For topography and keratoconus grading, the Keratoconus Severity Score was used18, which is based on average corneal power and corneal higher-order aberrations (expressed in µm as root-mean-square values). The Keratoconus Severity Score scale ranges from 0 (unaffected, normal topography), 1 (unaffected, atypical topography), 2 (suspect), 3 (mild keratoconus), 4 (moderate keratoconus) to 5 (severe keratoconus). Participant exclusion criteria were the following: previous eye surgery, any eye disease other than keratoconus, chronic use of topical medications or corneal opacities. Patients were asked to cease contact lens wear at least three weeks before examination.
AC C
EP
TE D
Three independent groups were recruited for this study. Patients with unremarkable slit lamp exam and normal topography (Keratoconus Severity Score 0: typical axial topography pattern, average corneal power ≤ 47.75 D and higher-order aberrations < 0.65 µm) in both eyes were recruited in group 1 (only one eye was randomly selected). Such strict topographical criteria were established to restrict the sample to patients without suspicious clinical and topographical findings and were adapted from Buhren et al19. Patients with biomicroscopy and topographical findings of manifest keratoconus (Keratoconus Severity Score 3 or greater: axial topography pattern consistent with keratoconus, may have positive slit-lamp findings, no corneal scarring, average corneal power > 49.00 D and higher-order aberrations > 1.50 µm) in at least one eye were considered keratoconic, and corneal curvature was used as a surrogate indicator of ectasia grade20: the eye with the lowest average corneal power was included in group 2 (better or less advanced keratoconic eyes) while the eye with the highest average corneal power was included in group 3 (worse or more advanced keratoconic eyes). No patient had exactly the same average corneal power in both eyes. Eyes in group 2 with insufficient topographical findings to meet diagnostic criteria for keratoconus (Keratoconus Severity Score 0-2) were considered subclinical cases and analyzed further as subgroup 2SC. The following corneal descriptors were obtained from the Pentacam’s software Belin-Ambrósio Enhanced Ectasia Display: central and thinnest pachymetry, front and back corneal elevation at the thinnest corneal point, pachymetric progression indices (minimum, average and maximum), normalized indices Df (deviation of normality of the front elevation), Db (deviation of normality of the back elevation), Dp (deviation of normality of pachymetric progression), Dt (deviation of normality of corneal thinnest point), Da (deviation of normality of relational thickness) and D
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
(overall deviation of normality), and the Ambrosio’s average and maximum relational thickness indices. The methodology for calculation has been disclosed by the manufacturer for some but not all of these indices6,11. Pachymetric progression refers to the percentage in corneal thickness increase along each meridian starting from the thinnest corneal point, and the minimum, average and maximum values are reported, whereas relational thickness indices (average and maximum) express the ratio of the thinnest pachymetry and the respective pachymetric progression. Upon request, the manufacturer informed that the deviation indices, average pachymetric progression and maximum relational thickness are classified by the software as normal (< 1.6 standard deviation –SD- from the population mean, shown in white), suspicious (≥ 1.6 SD and < 2.6 SD, highlighted in yellow) and pathologic (≥ 2.6 SD, highlighted in red), from the data reported by Ambrosio et al11, and this scheme was followed throughout this work. For the normalized indices, each observation was classified into one of the three categories, whereas for the pachymetric progression and relational thickness indices, each observation was categorized as normal or abnormal according to the cutoff values reported by Ambrosio et al11: minimum pachymetric progression ≥ 0.79, average pachymetric progression ≥ 1.06, maximum pachymetric progression ≥ 1.44, average relational thickness ≤ 424 and maximum relational thickness ≤ 339 were considered abnormal.
AC C
EP
TE D
Normality of the data was assessed by the Kolmogorov–Smirnov test, and parametric or nonparametric tests were subsequently used. Receiver operating characteristic (ROC) curves were used to calculate sensitivity, specificity and area under the curve (AROC) of each corneal index, as previously described4. Optimal cutoff points for each index were obtained from the ROC curves as those closest to the perfect classification point4. Logistic regression with forward stepwise inclusion (logistic function 1: logit = 0.467 x Db + 2.625 x Da – 4.871) and all variables entered at once (logistic function 2: logit = 0.333 x Df + 0.353 x Db – 0.466 x Dp + 0.077 x Dt + 3.149 x Da – 5.257) was employed to combine the 5 individual D indices, with a 0.5 cutoff. Statistical tests and analysis were performed with Prism 5 software (GraphPad Software, La Jolla, CA, USA) and SPSS 17 software (SPSS Inc., Chicago, IL, USA). Statistical significance was set at p < 0.05 and data are shown as mean±standard deviation unless otherwise stated. Data collection and sorting were done with the aid of Microsoft Excel 2010 software.
ACCEPTED MANUSCRIPT
Results Demographics and keratoconus grade distribution
Corneal descriptors in control and keratoconic eyes
SC
RI PT
Group 1 (control eyes) comprised 189 eyes (94 right and 95 left) from 189 patients with unremarkable topography (Keratoconus Severity Score 0) in both eyes. The less advanced eye (29 right and 26 left) of 55 keratoconus patients was included in group 2 whereas the worse eye (26 right and 29 left) was included in group 3. There was no significant difference in mean age between control (32.3±8.1 years, range 14-71) and keratoconic (32.5±11.7 years, range 14-65) patients, but gender distribution significantly differed (41% vs 60% males, p=0.01) between groups. According to Keratoconus Severity Score grading, group 2 included 21 (38%) grade 0 eyes, 7 (13%) grade 1 eyes, 9 (16%) grade 2 eyes, 15 (27%) grade 3 eyes, 1 (1%) grade 4 eye and 2 (4%) grade 5 eyes, whereas group 3 comprised 33 (60%) grade 3 eyes, 11 (20%) grade 4 eyes and 11 (20%) grade 5 eyes.
M AN U
Conventional corneal descriptors are summarized in Table 1 and depicted in Figure 1. Means for every Pentacam corneal descriptor in groups 2 and 3 were significantly different when compared with group 1 (Table 1, Figures 2 and 3). In group 2 eyes, every keratoconus index was significantly correlated with Keratoconus Severity Score grading (Spearman’s correlation, Table 1 for r values, p < 0.001). Truly subclinical keratoconus eyes (n=37 [67%], group 2 eyes with Keratoconus Severity Score 0-2) were analyzed further as subgroup 2SC (Table 1), and they also differed significantly in means with group 1 eyes for every Pentacam corneal descriptor analyzed.
TE D
Classification of control eyes according to their Pentacam corneal indices
AC C
EP
The classification results according to each keratoconus detection index are summarized in Table 2 and Figure 4. In group 1, 129 eyes (68.3%) had normal values for every D index, and of these, all but three (126 eyes, 66.7%) showed also normal values for maximum relational thickness and all but 21 (108 eyes, 57.1%) also exhibited normal average pachymetric progression values. When considering every deviation index and maximum relational thickness and average pachymetric progression altogether, 107 eyes (56.6%) showed normal values for every index. Regarding the pachymetric progression indices, only 128 eyes (67.7%) exhibited normal values in all of them, whereas for average and maximum relational thickness, 169 eyes (89.4%) had normal values in both indices. Only 99 eyes (52.4%) in group 1 did not show abnormal scores in each of the 11 indices reported by the Pentacam. Group 1 eyes with normal overall deviation values (n=152, 80.4%) did not significantly differ in higher-order aberrations with those with suspicious or abnormal values (0.250±0.072 vs 0.256±0.060 µm, p=0.66). Group 1 eyes with normal pachymetric progression deviation values (n=173, 91.5%) also did not differ in higherorder aberrations when compared with those group 1 eyes flagged as suspicious or abnormal (≥1.6 SD) according to this metric. Keratoconus detection by Pentacam corneal indices In group 3 (worse eyes), all observations had 7 or more abnormal values in the 11 keratoconus detection indices evaluated, whereas in group 2 (better eyes), 2 (3.6%) observations exhibited no abnormal indices and 7 (12.7%) observations had 2 or less abnormal indices (Figure 4). Considering a suspicious value (≥1.6 SD) as positive (Table 2 and Figure 2), the sensitivity of
ACCEPTED MANUSCRIPT
RI PT
front, back, pachymetric progression, corneal thickness, relational thickness and overall deviation indices for group 2 eyes was 70.9%, 69.1%, 78.2%, 58.2%, 81.8% and 92.7%, respectively, whereas for subclinical keratoconus eyes (subgroup 2SC) the sensitivity was 56.8%, 54.1%, 67.6%, 43.2%, 73.0% and 89.2%, respectively. Within group 2, true positive cases for every deviation index exhibited greater mean higher-order aberrations than false negative cases, and compared with group 1 eyes, mean higher-order aberrations were significantly increased in false negative eyes for all metrics except for the overall deviation index. Considering the previously reported cutoff values (Table 2 and Figure 3), the sensitivity of the minimum, average and maximum pachymetric progression and average and maximum relational thickness indices for group 2 eyes was 76.4%, 89.1%, 89.1%, 80.0% and 85.5%, respectively, whereas for subclinical keratoconus eyes (subgroup 2SC) was 64.9%, 83.8%, 83.8%, 70.3% and 78.4%, respectively.
AC C
EP
TE D
M AN U
SC
A logistic regression approach with the 5 individual deviation indices selected the combination of back and relational thickness indices as the best possible. This diagnostic function predicted 186 eyes in group 1 (98.4%) as healthy and 44 eyes in group 2 (80.0%) as keratoconic, whereas a logistic regression approach with the 5 individual deviation indices entered at once predicted 187 eyes in group 1 (98.9%) as healthy and 44 eyes in group 2 (80.0%) as keratoconic. In comparison, the Pentacam’s overall deviation index with a ≥2.6 SD cutoff marked 186 eyes in group 1 as healthy and 44 eyes in group 2 as keratoconic, and the classification matched that of the first logistic function except for one case in the keratoconus group. An independent ROC analysis was performed between groups 1 and 2 in order to validate the previously published cutoff values for each Pentacam corneal index, and the sensitivity and specificity are summarized in Table 3 along with optimal cutoff values derived from these samples.
ACCEPTED MANUSCRIPT
Discussion
RI PT
Pentacam Scheimpflug tomography provides specific corneal analysis indices designed to diagnose ectatic changes. In this work we explored the actual performance of these descriptors in a large sample of topographically-normal patients and in keratoconus cases to assess which indices are more adequate for diagnosing this disorder at its initial stage. The results altogether show that Scheimpflug tomography allows the clinician to find early ectatic changes, and at the same time, that there is still a considerable fraction of subclinical keratoconus cases that can go undetected even by this method.
M AN U
SC
In agreement with previous reports that utilized diverse criteria for defining keratoconus6,7,9–16, we observed that all Pentacam indices differed significantly between healthy eyes and the better eyes of keratoconus patients diagnosed by the findings in their more advanced fellow eyes. In this keratoconic sample, two out of three eyes (n=37, 67%) could be considered as subclinical cases, that is, with insufficient topography findings to meet the Keratoconus Severity Score criteria for manifest disease, and most of this subset (21 eyes, 57%) had unremarkable Placido topography. Even this stricter subset had reduced corneal thickness, increased anterior surface irregularity and showed significant differences in every Pentacam index, suggesting that at least slight changes in corneal shape were present in these keratoconic eyes with subclinical disease, as others have shown2,19. There was however considerable overlap in Pentacam indices between group 1 eyes and keratoconic eyes (Figures 2 and 3), which affected their diagnostic performance.
AC C
EP
TE D
Regarding the standardized indices, which are the ones that offer the most straightforward interpretation in the Pentacam report, the diagnostic yield varied markedly between each metric. Although increased posterior elevation has been proposed as a sensitive marker for keratoconus10,21, back corneal surface deviation index had poor performance for subclinical keratoconus detection and was not clinically superior to anterior corneal surface deviation, as both indices flagged only about one half of the subclinical keratoconus eyes as abnormal (Table 2). Our findings suggest that detectable back corneal surface abnormalities do not necessarily arise before anterior changes. This result, however, differs with that reported by de Sanctis et al10,21, which was still low (68-73% sensitivity), and probably derives from the more stringent diagnostic criteria that we employed for defining subclinical disease. Supporting this idea, the sensitivity of posterior corneal elevation in our sample was about 67% when we considered the better fellow eyes of manifest keratoconus cases (group 2). Corneal thickness deviation had the lowest sensitivity (43%) among the studied indices, and this result simply reflects the significant overlap in pachymetric values between healthy eyes and truly subclinical keratoconus cases14,16. In contrast to front and back elevation, focalized corneal thinning seems to typify some of the earliest changes in keratoconic eyes. The Pentacam software provides several metrics designed to detect these ectatic signs: pachymetric progression and Ambrosio’s relational thickness indices6,11, along with standardized versions to ease interpretation. According to the reported cutoff points11, average and maximum pachymetric progression were the single metrics with the highest sensitivity (83.8%) for subclinical keratoconus, and maximum relational thickness followed closely in our sample. Nevertheless their performance was not as good as initially reported and probably reflects the fact that we tested them on a subclinical keratoconus sample instead of patients with bilateral disease11. Remarkably, the respective standardized indices exhibited lower sensitivity than their unstandardized counterparts, and this was especially true for standardized pachymetric progression, which only detected about two thirds of the subclinical
ACCEPTED MANUSCRIPT
RI PT
keratoconus eyes in our sample. An independent ROC analysis (Table 3) yielded optimal cutoff values that were similar to those initially reported, albeit slightly displaced towards the normal mean for every metric, evidencing the different composition of the keratoconic sample. The optimized cutoff value of 1.26) conveyed comparable levels of detection, but this was not true for the standardized pachymetric progression index. It should be noted that these findings still require independent validation as they might reflect the specific composition of our sample.
EP
TE D
M AN U
SC
As no specific cutoff value for each individual deviation index completely separated both groups, a multivariate approach that combined several indices might yield better diagnostic performance. The Pentacam software provides such metric, the overall deviation index, although the exact statistical method underlying its calculation was not disclosed by the manufacturer and its actual performance has not been independently validated. In our sample this index correctly identified 89.2% of the subclinical keratoconus cases, providing one of the highest sensitivities among the Pentacam’s metrics. We conducted our own analysis of combinations of the deviation indices, and either by selecting the best possibility with the fewest indices (back and relational thickness deviation) or the best possibility with every index we obtained almost identical results to those of the overall deviation index with a ≥2.6 SD cutoff, which is the actual value reported as pathologic by the Pentacam software. We infer from this analysis that the unknown algorithm built in the software already provides optimal results, and more importantly, that there is a limit to the diagnostic capacity of the slight changes in corneal shape detected by Scheimpflug tomography. We chose to consider a suspicious value (≥1.6 SD) as positive in order to maximize sensitivity while sacrificing specificity; in other words, because we believe that in the preoperative evaluation of refractive surgery candidates it is preferable to falsely flag a cornea as ectatic than to miss a subclinical keratoconus case. The Pentacam’s suspicious cutoff value for the overall deviation index was validated by our independent ROC analysis, which suggested >1.61 as optimal for this particular keratoconic sample. This approach, however, predicted about one every five eyes in the control group to be potentially keratoconic, an excessively high false positive rate (specificity 82.3%). In contrast, modified cutoff values for standardized and maximum relational thickness offered comparable sensitivity and improved specificity in our sample (Table 3), and perhaps clinicians should rely more on these indices to reduce the false positive rate. These findings nevertheless await independent validation.
AC C
There is one considerable limitation in this study, which is the composition of the control group. As we attempted to reproduce the clinical setting in which Pentacam analysis takes place, only clinical and topographical data gathered at the same visit were considered. In this way we followed similar criteria to other reports, which include standard Keratoconus Severity Score criteria for unremarkable slit lamp exam and corneal topography in both eyes. Nonetheless a significant proportion of subclinical keratoconus cases have normal topography (defined as Keratoconus Severity Score 0), as evidenced by the group 2 keratoconic eyes, and considering the large sample size of the control group (n=189) and the keratoconus prevalence reported elsewhere13, it is expected that a few subclinical keratoconus cases would be artificially included in this group and therefore specificity data from it derived could be inaccurate. This seems to be the case for at least three observations that showed markedly abnormal overall deviation values. The remaining 34 cases with a suspicious overall deviation value probably comprise false
ACCEPTED MANUSCRIPT
positives that will never develop ectatic changes and some early keratoconus cases, but prospective studies are required to properly address this matter.
AC C
EP
TE D
M AN U
SC
RI PT
In conclusion, we found that Pentacam Scheimpflug tomography analysis can detect most but not all subclinical keratoconus cases with insufficient anterior surface findings to be considered as keratoconic according to the Keratoconus Severity Score scale. As provided by the software, the multivariate overall deviation metric was most appropriate for this task, whereas the performance of the individual indices fell considerably behind unless modified cutoff values were used. We also observed that when evaluating apparently healthy subjects with unremarkable topography, there are a sizable number of cases that are flagged as suspects by the Pentacam analysis, and that additional testing should perhaps be applied to these patients. There are other keratoconus detection tools, such as corneal biomechanical testing4,22 and epithelial thickness analysis5,23, which rely on different concepts and could be used complementarily with Scheimpflug tomography.
ACCEPTED MANUSCRIPT
Acknowledgements
AC C
EP
TE D
M AN U
SC
RI PT
a) Funding/Support: None b) Financial Disclosures: None c) Author Contributions: design (FFB, TP, JGG) and conduct of the study (PRV, JDG, NM and MD); collection (PRV, JDG, NM and MD), analysis (JGG), and interpretation of the data (PRV, JDG and JGG); and preparation (PRV, JDG and JGG), review (NM), and approval of the manuscript (FFB and TP). d) Other Acknowledgments: None
ACCEPTED MANUSCRIPT
References
6.
7.
8.
9.
10.
11. 12. 13.
14.
15.
16.
RI PT
SC
5.
M AN U
4.
TE D
3.
EP
2.
Klein SR, Epstein RJ, Randleman JB, Stulting RD. Corneal ectasia after laser in situ keratomileusis in patients without apparent preoperative risk factors. Cornea. 2006;25(4):388–403. Bühren J, Schäffeler T, Kohnen T. Preoperative topographic characteristics of eyes that developed postoperative LASIK keratectasia. J Refract Surg. 2013;29(8):540–9. Belin MW, Ambrósio R. Scheimpflug imaging for keratoconus and ectatic disease. Indian J Ophthalmol. 2013;61(8):401–6. Galletti JG, Pförtner T, Fuentes Bonthoux F. Improved keratoconus detection by ocular response analyzer testing after consideration of corneal thickness as a confounding factor. J Refract Surg. 2012;28(3):202–8. Reinstein DZ, Archer TJ, Gobbe M. Stability of LASIK in topographically suspect keratoconus confirmed non-keratoconic by Artemis VHF digital ultrasound epithelial thickness mapping: 1-year follow-up. J Refract Surg. 2009;25(7):569–77. Ambrósio R, Alonso RS, Luz A, Coca Velarde LG. Corneal-thickness spatial profile and corneal-volume distribution: tomographic indices to detect keratoconus. J Cataract Refract Surg. 2006;32(11):1851–9. Smadja D, Santhiago MR, Mello GR, Krueger RR, Colin J, Touboul D. Influence of the reference surface shape for discriminating between normal corneas, subclinical keratoconus, and keratoconus. J Refract Surg. 2013;29(4):274–81. Piñero DP, Alió JL, Alesón A, Escaf Vergara M, Miranda M. Corneal volume, pachymetry, and correlation of anterior and posterior corneal shape in subclinical and different stages of clinical keratoconus. J Cataract Refract Surg. 2010;36(5):814–25. Nilforoushan M-R, Speaker M, Marmor M, et al. Comparative evaluation of refractive surgery candidates with Placido topography, Orbscan II, Pentacam, and wavefront analysis. J Cataract Refract Surg. 2008;34(4):623–31. De Sanctis U, Loiacono C, Richiardi L, Turco D, Mutani B, Grignolo FM. Sensitivity and specificity of posterior corneal elevation measured by Pentacam in discriminating keratoconus/subclinical keratoconus. Ophthalmology. 2008;115(9):1534–9. Ambrósio R, Caiado ALC, Guerra FP, et al. Novel pachymetric parameters based on corneal tomography for diagnosing keratoconus. J Refract Surg. 2011;27(10):753–8. Miháltz K, Kovács I, Takács A, Nagy ZZ. Evaluation of keratometric, pachymetric, and elevation parameters of keratoconic corneas with pentacam. Cornea. 2009;28(9):976–80. Serdarogullari H, Tetikoglu M, Karahan H, Altin F, Elcioglu M. Prevalence of keratoconus and subclinical keratoconus in subjects with astigmatism using pentacam derived parameters. J Ophthalmic Vis Res. 2013;8(3):213–9. Uçakhan ÖÖ, Cetinkor V, Özkan M, Kanpolat A. Evaluation of Scheimpflug imaging parameters in subclinical keratoconus, keratoconus, and normal eyes. J Cataract Refract Surg. 2011;37(6):1116–24. Ahmadi Hosseini SM, Abolbashari F, Niyazmand H, Sedaghat MR. Efficacy of corneal tomography parameters and biomechanical characteristic in keratoconus detection. Cont Lens Anterior Eye. 2014;37(1):26–30. Muftuoglu O, Ayar O, Ozulken K, Ozyol E, Akıncı A. Posterior corneal elevation and back difference corneal elevation in diagnosing forme fruste keratoconus in the fellow eyes of unilateral keratoconus patients. J Cataract Refract Surg. 2013;39(9):1348–57.
AC C
1.
ACCEPTED MANUSCRIPT
23.
RI PT
22.
SC
21.
M AN U
20.
TE D
19.
EP
18.
Ambrósio R, Nogueira LP, Caldas DL, et al. Evaluation of corneal shape and biomechanics before LASIK. Int Ophthalmol Clin. 2011;51(2):11–38. McMahon TT, Szczotka-Flynn L, Barr JT, et al. A new method for grading the severity of keratoconus: the Keratoconus Severity Score (KSS). Cornea. 2006;25(7):794–800. Buhren J, Kuhne C, Kohnen T. Defining subclinical keratoconus using corneal firstsurface higher-order aberrations. Am J Ophthalmol. 2007;143(3):381–9. Zadnik K, Barr JT, Edrington TB, et al. Baseline findings in the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) Study. Invest Ophthalmol Vis Sci. 1998;39(13):2537–46. De Sanctis U, Aragno V, Dalmasso P, Brusasco L, Grignolo F. Diagnosis of subclinical keratoconus using posterior elevation measured with 2 different methods. Cornea. 2013;32(7):911–5. Ruiseñor Vázquez PR, Delrivo M, Fuentes Bonthoux F, Pförtner T, Galletti JG. Combining Ocular Response Analyzer Metrics for Corneal Biomechanical Diagnosis. J Refract Surg. 2013;29(9):596-602. Rocha KM, Perez-Straziota CE, Perez-Straziota E, Stulting RD, Randleman JB. SD-OCT analysis of regional epithelial thickness profiles in keratoconus, postoperative corneal ectasia, and normal eyes. J Refract Surg. 2013;29(3):173–9.
AC C
17.
ACCEPTED MANUSCRIPT
Figure Captions
RI PT
Figure 1. Conventional corneal descriptors in control and keratoconus eyes. Box-whisker plots of average corneal power (ACP, expressed in diopters –D-), corneal higher-order aberrations (HOA), central corneal thickness (CCT), thinnest pachymetry (Thinnest pach.) and front and back elevation at the thinnest corneal location in control eyes (group 1, empty boxes) and better (group 2, light-gray boxes) and worse (group 3, dark-gray boxes) eyes of keratoconus patients. * indicates a statistically-significant difference in mean compared with the control group (analysis of variance with Dunnett’s post hoc test or non-parametric equivalent).
M AN U
SC
Figure 2. Standardized deviation indices in control and keratoconus eyes. Box-whisker plots of deviation of front (Df) and back (Db) elevation, pachymetric progression (Dp), corneal thickness (Dt), relational thickness (Da) and combined index (D) in control eyes (group 1, empty boxes) and better (group 2, light-gray boxes) and worse (group 3, dark-gray boxes) eyes of keratoconus patients. The reported suspect (≥1.6 standard deviation –SD=) and abnormal (≥2.6 SD) cutoff values are plotted for reference. * indicates a statistically-significant difference in mean compared with the control group (analysis of variance with Dunnett’s post hoc test or nonparametric equivalent).
TE D
Figure 3. Pachymetric progression and relational thickness indices in control and keratoconus eyes. Box-whisker plots of minimum (PPImin), average (PPIave) and maximum (PPImax) pachymetric progression indices and average (ARTave) and maximum (ARTmax) Ambrosio’s relational thickness indices in control eyes (group 1, empty boxes) and better (group 2, light-gray boxes) and worse (group 3, dark-gray boxes) eyes of keratoconus patients. The reported cutoff values are plotted for reference. * indicates a statistically-significant difference in mean compared with the control group (analysis of variance with Dunnett’s post hoc test or nonparametric equivalent).
AC C
EP
Figure 4. Distribution of Pentacam’s keratoconus detection indices in control and keratoconus eyes. Frequency distribution (as percentage of total cases) of the number of abnormal indices (6 deviation indices, 3 pachymetric progression indices and 2 relational thickness indices) in control eyes (group 1), subclinical keratoconus eyes (subgroup 2SC) and better (group 2) and worse (group 3) eyes of keratoconus patients.
ACCEPTED MANUSCRIPT
Pentacam’s keratoconus indices
PPImax Df
Db Dp Dt
Da D
-0.21±0.90 (-1.49-4.05) 0.44±0.92 (-1.99-3.63) 0.33±0.96 (-2.16-3.09)
2.00±2.31# (-1.30-7.58) 3.49±3.24# (-1.05-13.94) 1.64±1.31# (-0.91-4.84)
4.15±4.44# (-1.30-23.22) , r=0.808 5.97±5.09# (-1.05-18.15), r=0.772 2.31±1.85# (-0.91-9.28), r=0.479
0.39±0.79 (-2.62-2.29) 1.03±0.70 (-0.87-4.86) 555±98 (333-893) 445±87 (237-770)
2.08±0.94# (-1.20-3.74) 4.26±3.03# (0.55-16.55) 381±125# (160-744) 269±101# (97-620)
2.48±1.00# (-1.20-3.99), r=0.664 6.08±3.98# (0.55-16.55), r=0.757 318±142# (112-744), r=-0.757 223±109# (78-620), r=-0.729
AC C
ARTave (µm) ARTmax (µm)
Keratoconus patients (n=55) Better keratoconus eyes Group 2 (n=55)
Worse keratoconus eyes Group 3 (n=55) 51.53±7.48# (42.23-77.24) 3.261±5.918# (0.373-36.840) 479±76# (321-940) 443±45# (298-550), r=0.415 26±15# (-11-72), r=0.676 54±26# (-1-130), r=0.667 1.90±0.86# (0.59-4.28), r=0.695 2.53±0.99# (1.03-5.61), r=0.741 3.70±1.53# (1.41-8.35), r=0.670 11.68±8.25# (-1.86-35.29), r=0.821 9.63±6.98# (-1.03-29.83),r=0.854 11.01±6.62# (0.84-31.85), r=0.736 3.50±2.06# (-0.35-12.46), r=0.419 3.21±0.96# (1.00-8.28), r=0.636 9.81±4.70# (1.42-22.92), r=0.833 208±99# (53-518), r=-0.704 145±73# (42-379), r=-0.662
SC
RI PT
46.03±3.50# (40.76-56.56) 1.153±0.947# (0.183-5.596) 491±39# (407-587), r=0.364 474±43# (376-572), r=0.474 16±12# (-3-43), r=0.779 36±27# (0-90), r=0.757 1.34±0.70# (0.46-3.48), r=0.722 1.79±0.77# (0.75-3.67), r=0.773 2.61±1.17# (0.90-5.35), r=0.743 5.43±5.18# (-1.19-21.97) , r=0.850
M AN U
PPIave
Subclinical keratoconus eyes Group 2SC (n=37) 44.71±2.19 (40.76-48.90) 0.669±0.350# (0.183-1.472) 502±37# (412-587) 489±37# (409-572) 10±8# (-3-36) 24±21# (0-85) 1.02±0.41# (0.46-2.25) 1.42±0.48# (0.75-2.97) 2.06±0.81# (0.90-4.90), r=0.464 2.71±2.48# (-1.19-8.67)
TE D
ACP (D) HOA (µm) CCT (µm) TP (µm) Front elevation at thinnest location Back elevation at thinnest location PPImin
Control patients Group 1 (n=189) 44.48±1.48 (40.83-47.48) 0.251±0.070 (0.145-0.497) 532±33 (449-628) 528±32 (448-625) 3±2 (-3-14) 4±5 (-7-37) 0.70±0.13 (0.37-1.16) 0.97±0.14 (0.61-1.44) 1.22±0.20 (0.73-2.02) 0.44±1.07 (-1.79-4.13)
EP
Conventional corneal descriptors
Table 1 – Topography & Pentacam corneal descriptors in control and keratoconic eyes
Data is shown as mean±standard deviation (minimum-maximum). 2SC = group 2 eyes with Keratoconus Severity Score 0; ACP = average corneal power; HOA = higherorder aberrations; CCT = central corneal thickness; TP = thinnest pachymetry; PPI = pachymetric progression indices (min = minimum, ave = average, max = maximum); Df = deviation of front elevation; Db = deviation of back elevation; Dp = deviation of pachymetric progression; Dt = deviation of corneal thickness; Da = deviation of relational thickness; D = overall deviation; ART = Ambrosio’s relational thickness indices (ave = average, max = maximum). # Statistically-significant difference with group 1 eyes.
ACCEPTED MANUSCRIPT
Table 2 – Classification table for Pentacam keratoconus indices Control eyes Group 1 (n=189) Abnorma Normal Suspect l (1.29
67.6
76.9
>1.00
89.2
90.3
>1.26
89.2
82.3
>1.61
73.0
73.7
78.4
82.8
86.5
87.1
88.9
78.4
90.5
86.5
AROC 0.86 (0.80-0.93) 0.86 (0.80-0.93) 0.91 (0.85-0.96) 0.85 (0.78-0.91) 0.96 (0.92-0.99) 0.95 (0.92-0.99) 0.85 (0.78-0.93) 0.90 (0.85-0.96) 0.95 (0.91-0.99) 0.92 (0.87-0.97) 0.95 (0.91-0.99)
>0.76
>1.09
>1.41 0.77
85.5
83.3
>1.29
72.7
82.3
>1.21
SC
Db
0.80 (0.71-0.89) 0.80 (0.71-0.89) 0.86 (0.78-0.94) 0.79 (0.71-0.88) 0.94 (0.88-0.99) 0.93 (0.88-0.98) 0.79 (0.68-0.89) 0.86 (0.78-0.94) 0.92 (0.86-0.98) 0.88 (0.80-0.95) 0.93 (0.87-0.98)
Spec
92.7
90.3
>1.26
85.5
95.7
>2.17
74.5
89.8
>0.85
85.5
82.2
>1.09
89.1
90.3
>1.46
88.9
85.5
