PURPOSE

To assess the reliability of successive Corvis ST® measurements (CST, Oculus, Wetzlar, Germany) in keratoconus (KC) ≥ 2 years after accelerated corneal crosslinking (9 mW/cm, 10 min, 5.4 J/cm) compared to untreated KC corneas.

METHODS

Three successive CST measurements per eye were performed in ≥ 2 years after CXL (CXLG, n = 20 corneas of 16 patients) and a control group consisting of non-operated, ABC-stage-matched KC corneas according to Belin's ABCD KC grading (controls, n = 20 corneas, 20 patients). Main outcome measures included maximal keratometry (Kmax), the Belin/Ambrósio-Enhanced-Ectasia-Deviation-Index BAD-D; the biomechanical parameters A1 velocity, deformation amplitude (DA) ratio 2 mm, Ambrósio relational thickness to the horizontal profile (ARTh), integrated radius, stiffness parameter A1 (SP-A1), and the Corvis Biomechanical Factor (CBiF, the linearized term of the Corvis Biomechanical Index). Mean values, standard deviations, and Cronbach's alpha (CA) were calculated.

RESULTS

Both groups were tomographically comparable (BAD: 11.5 ± 4.7|11.2 ± 3.6, p = 0.682, Kmax: 60.5 ± 7.2|60.7 ± 7.7, p = 0.868 for controls|CXLG, paired t-test). A1 velocity (mean ± SD: 0.176 ± 0.02|0.183 ± 0.02, p = 0.090, CA: 0.960|0.960), DA ratio 2 mm (6.04 ± 1.13|6.14 ± 1.03, p = 0.490, CA: 0.967|0.967), integrated radius (12.08 ± 2.5|12.42 ± 1.9, p = 0.450, CA: 0.976|0.976), and CBiF (4.62 ± 0.6|4.62 ± 0.4, p = 0.830, CA: 0.965|0.965) were also comparable (controls|CXLG). ARTh was significantly higher in controls (177.1 ± 59, CA: 0.993) than after CXL (155.21 ± 65, p = 0.0062, CA: 0.993) and SP-A1 was significantly higher after CXL (59.2 ± 13, CA: 0.912) than in controls (52.2 ± 16, p = 0.0018, CA: 0.912).

CONCLUSION

ARTh and SP-A1 differed significantly between controls and CXLG. Biomechanical measurements were generally of excellent reliability in both groups. CXL seems to affect biomechanical measurements of human corneas over more than 2 years.