Topography, tomography and biomechanical assessment for diagnosing forme fruste keratoconus

Session Details

Session Title: Presented Poster Session 19: Phakic IOLs 2

Session Date/Time: Tuesday 16/09/2014 | 09:30-11:00

Paper Time: 10:45

Venue: Pod 1 (Poster Village)

First Author: : J.Steinberg GERMANY

Co Author(s): :    K. Lücke   A. Frings   M. Casagrande   V. Druchkiv   T. Katz   S. Linke

Abstract Details

Purpose:

To discriminate between normal eyes (NE) and forme fruste keratoconus eyes (ffKK) by applying Scheimpflug-technique based corneal topography, tomography (both Pentacam®) and biomechanical analyses (CorvisST®).

Setting:

University Medical Centre Hamburg-Eppendorf, Germany.

Methods:

Retrospective, cross-sectional data analysis. Pentacam® and CorvisST® analyses of eighteen forme fruste keratoconus eyes (ffKK) of eighteen subjects with unilateral keratoconus and eighteen eyes of eighteen subjects with bilateral normal eyes (NE) were reviewed. All included eyes displayed KISA% <100. Because of the biomechanical analyses, both groups were matched with respect to the intraocular pressure. Receiver operating characteristics (ROC) and logit estimation were performed to develop a discrimination index (DI).

Results:

Despite including far more than 100 topographic, tomographic and biomechanic parameters, only one topographic (paracentric sagittal radius [PS]) and one tomographic parameter (vertical deviation of the thinnest point [BAD_dy]) displayed significant differences between both groups. None of the included biomechanical parameters reached a statistical significant level. Combining PS and BAD_dy resulted in a DI with an area under the curve (AUC) of 0.92 (corresponding sensitivity/specificity both 0.83).

Conclusions:

Combining Scheimpflug-based-analyses of corneal topography, tomography enables discrimination of normal and ffKK eyes with high sensitivity and specificity. In our study, the addition of biomechanical parameter did not improve the diagnosis of ffKK. However, because of the potential benefit of this new technique, multicenter studies including more subjects should be conducted to evaluate our results.