Evaluation of the Robustness of Current Quantitative Criteria for Keratoconus Progression and Corneal Cross-linking.

PURPOSE: To evaluate the robustness of numerical indications of corneal cross-linking by evaluating the inter-session reliability of recently used parameters of progression in keratoconus.

METHODS

In this observational study, 100 eyes of 100 patients with keratoconus underwent analysis with the Sirius Scheimpflug Topography System (CSO, Firenze, Italy) by a single user. Two sessions (three scans per session) were performed 2 to 4 weeks apart. Cases having poor quality scans, recent visual deterioration, and ocular surgery were excluded. The corneal variables measured were thinnest point and apex characteristics (pachymetry, polar coordinates), volume, central simulated keratometry (flat, steep, average, astigmatism), central corneal thickness, maximum elevation (anterior, posterior), symmetry indices (front, back), and Baiocchi-Calossi-Versaci index. Intra-session (first session, all three measurements) and inter-session (one measurement from each session, selected randomly) statistics were computed.

RESULTS

There were no significant differences in the six measurements for all variables (P > .05, analysis of variance). The intra-session and inter-session intraclass correlations were high (0.937 to 0.997) and within-subject standard deviations (Sw) were satisfactory (< 5 µm for thickness, < 0.50 diopters [D] for curvature, and < 0.11 mm, < 5° for polar coordinates). Inter-session repeatability (2.77 × Sw) was found to be satisfactory when compared to previously used central keratometric (> 1.00 D increase) or pachymetric (> 2%/> 5% decrease) guidelines to define progression. Apex power repeatability (1.24 D) was found to be poorer than previously used recommendations (1.00 D change).

CONCLUSIONS

The criteria used to define keratoconus progression were satisfactory when compared to inter-session reliability of corneal parameters. Because higher variability was noted at apex, its curvature repeatability cut-off may be raised to 1.25 D for identifying progressive keratoconus. [J Refract Surg. 2016;32(7):465-472.].