To evaluate the differences in central and non-central keratoconus (based on cone location), and their effect on the objective screening thresholds for keratoconus.
This comparative case series was performed at tertiary care cornea and refractive surgery service. Three groups were made: KC apex within central 2 mm (central keratoconus, n = 50), apex outside central 2mm (non-central keratoconus, n = 50) and normal controls (n = 100, with 50 cases each with apex within and outside central 2 mm). All cases underwent clinical evaluation and corneal topography (CSO, Sirius, Italy). Apex keratometry (ApexK), simulated keratometry at 3 mm (SimK), central corneal thickness (CCT) and minimum corneal thickness (MCT), anterior corneal higher-order aberrations root mean square (HOARMS), and Zernike's coefficients up to fourth order at different zones were measured.
In spite of the keratoconic groups having comparable ApexK (p > 0.05), central keratoconus had higher SimK and thinner CCT and MCT (p < 0.001). HOARMS was significantly more for central keratoconus at 3 mm zones. These findings had moderate to large effect size (Cohen's d). Receiver operating curve analysis was carried out to compare central keratoconus and non-central keratoconus with control group. ApexK and HOARMS had best discriminative parameters. Using single parametric suspicion cut-offs of 'either SimK steep >47.2 D or CCT < 491.6 μ' had a good sensitivity (0.98) for central keratoconus, but not for non-central keratoconus (0.80). Changing this cut-off to 'either SimK steep K ≥ 45.8 D or CCT ≤ 503 μ' gave a sensitivity and specificity of 0.95 and 0.87 for non-central keratoconus and 0.99 and 0.87 for central keratoconus.
Non-central keratoconus has lesser effect on SimK, pachymetry and smaller-aperture HOARMS. Using 'SimK steep >47.2 D or CCT < 491.6 μ' may miss timely referral for topography in many of these cases. Using more stringent criteria of SimK steep K ≥ 45.8 D or CCT ≤ 503 μ to get a corneal topography done to rule out keratoconus is recommended, especially in cohorts with higher risk.