INTRAOCULAR LENS CALCULATION IN KERATOCONUS PATIENTS – WHICH FORMULA TO USE? WHICH MEASURE OF KERATOMETRIC POWER?

Intraocular lens calculation in keratoconus patients is challenging due to an abnormal ratio of the anterior to posterior corneal power, inaccurate estimations of the effective lens position due to erroneous corneal power measurements, and variable distribution of the corneal curvature along non-orthogonal meridians. Modern formulas were optimized to partially account for these errors. We aim to assess its accuracy in keratoconus patients, and to explore potential advantages of using tomography-derived queratometry and more advanced estimations of corneal power: true net power (TNP) and total corneal refractive power (TCRP).

Refractive Surgery, Ophthalmology Department – Central Lisbon Universitary Hospital Center, Lisbon

Retrospective case series of stage 2 and 3 keratoconus patients who underwent uneventful cataract surgery with in-the-bag implantation of an acrylic hydrophobic spherical IOL. Using ULIB-optimised constants, retrospective constant optimization and measurements from optical biometry, the keratoconus modification of the Kane and Barrett formulas were used to calculate the prediction error for each eye. The same prediction errors using those formulas were calculated using 3 other measures of keratometric power derived from corneal tomography (Pentacam): anterior keratometry, TNP and TCRP (4mm pupil-centered ring). Main outcome measures included mean prediction error (ME), mean (MAE) and median (MedAE) absolute errors in D, and the percentage of eyes within ±0.25, 0.50, 0.75 and 1.00D of the intended refractive target.

A total of 20 eyes from 13 patients were included for analysis. There was a significant difference (p<0.001) between the mean keratometry measured by the Lenstar (53.13±5.72), Pentacam (50.36±2.37), TNP (46.5±1.82) and TCRP (47.68±1.94). The order rank of lowest to highest MAE was: Barrett-Lenstar (1.302), Kane-Pentacam (1.331), Kane-Lenstar (1.338), Kane-TCRP (1.367), Kane-TNP (1.392), Barrett-Pentacam keratometry (1.622), Barrett-TNP (1.968) and Barrett-TCRP (2.038). Only Kane with Lenstar and Pentacam keratometry had standard deviation (SD) errors less than 1; Barrett with Pentacam, TNP and TCRP had the highest SD. Considering MedAE, Barrett-Pentacam (0.711), Kane-Pentacam (1.018), Barrett-Lenstar (1.136) and Kane-Lenstar (1.264) had the lowest values. Overall, refractive accuracy was poor, with only Barrett-Lenstar keratometry achieving more than half (55%, n=11) of patients within ±1.00D of the predicted refractive target.

In a moderate to severe group of keratoconus patients, IOL calculation is challenging, even when using modern formulas optimized for this corneal ectatic disease. When inputting biometric measurements for the keratoconus-modified Kane and Barrett formulas, anterior keratometry derived from both optical biometry and tomography provide acceptable results. In these patients, TNP and TCRP underestimate corneal power, lead to large myopic refractive surprises and therefore are not recommended.