This study compared the accuracy and repeatability of Fourier-domain optical coherence tomography (FD-OCT) with both ultrasonic pachymetry (USP) and a rotating Scheimpflug camera for measuring central corneal thickness (CCT). Measurements made with all three methods in normal subjects indicate that while there are significant correlations between all of them, FD-OCT significantly underestimates CCT compared to the other devices. The authors conclude that the RTVue-100 FD-OCT (Optovue, Inc., Fremont, Calif.) used in the study may be a useful alternative for measuring CCT but its measurements are not directly interchangeable in clinical practice with those made with the other two devices despite their high correlation.
The same examiner measured CCT in 30 subjects (30 normal corneas) using the RTVue-100 FD-OCT with an anterior segment adaptor, a Pentacam rotating Scheimpflug camera and SP-2000 USP. Two examiners obtained one FD-OCT measurement from 10 eyes of five subjects to assess interexaminer reproducibility.
Mean CCT (± SD) measured by FD-OCT, USP and the Pentacam were 530 ± 33, 544 ± 34 and 552 ± 35 mm, respectively. Significant correlations were found between FD-OCT and USP (P < 0.0001), FD-OCT and Pentacam (P < 0.0001) and USP and Pentacam (P < 0.0001). Pairwise comparisons showed that CCT measurement with FD-OCT was significantly thinner than with the other two methods (P < 0.001 for all comparisons). Analysis of intraexaminer repeatability demonstrated that the intraclass correlation coefficients ranged between 0.97 and 0.98. There was high repeatability of the CCT measurements with all methods.
The authors say that overestimation of CCT could lead to serious complications, especially in patients undergoing refractive surgery, whereas underestimation could lead to exclusion of otherwise eligible patients from refractive procedures. They say that FD-OCT is appropriate for examining corneas to obtain greater details, such as for assessing eyes before and after refractive surgery. However, its relatively shallow imaging depth and narrow width (maximum span, 6 mm) make it unsuitable for biometry of the entire cornea and anterior chamber.
They conclude that the differences in measurements result from substantially different operating principles. With USP, unlike with OCT, the ultrasound contact probe displaces the tear film and compresses the corneal surface, with the exact location of the posterior corneal reflection unknown, that is, possibly between Descemet membrane and the anterior chamber. In addition, misplacement of the probe might result in considerable measurement errors, although five USP measurements were taken and averaged. While the operating principle of the Pentacam depends on the reflectivity of light beams and differs substantially from OCT systems, the higher resolving power of FD-OCT may largely contribute to better corneal edge detection, resulting in underestimation.