INTRODUCTION

Cataract surgery, especially when combined with the correction of presbyopia, has also become a refractive procedure.

It is known that the presence of residual spherical or cylindrical refractive errors after the implantation of multifocal lenses significantly degrades visual quality and is still the main cause of dissatisfaction in these patients^1,2.

The minimum astigmatism that should be corrected during cataract surgery is controversial, since its influence on vision varies with individual factors, such as pupillary diameter, high order aberrations (HOAs), and the ability to adapt to patterns of aberrations. However, it has been suggested that the correction of astigmatism greater than 0.50 D improves the visual results³.

In a recently published series of 13,012 eyes of candidate patients for cataract surgery at Hospital da Luz, Lisbon, it was demonstrated that in this population the mean astigmatism was 1.08 ± 0.84 D, with the prevalence of astigmatism being greater than 1.00 D in 43.5% (Figure 1)⁴.

It should be noted that this prevalence of astigmatism higher than 1.00 D is greater than that found in most series published in European populations, namely Spain (34.8%) and Germany (36%)^5,6. Thus, it is essential to consider this correction in any presbyopia surgery.

Among the various strategies for the correction of astigmatism during cataract surgery, toric intraocular lenses offer the possibility of correcting a wide range of spherical and cylindrical refractive errors with high precision, predictability and excellent visual results7,8.

Multifocal or extended depth of focus toric IOLs allow for good visual acuity and independence of spectacles at all distances and are the option to consider in presbyopia surgery when dealing with patients with corneal astigmatism as any residual astigmatism greater than 0.75 D in the presence of these lenses may lead to worse results9.

The process of selecting an IOL for correcting presbyopia is time consuming, complex and susceptible to several errors in any of its stages.

It should start by accurately measuring astigmatism, preferably with different devices so that results can be compared, including the calculation of the spherical and cylindrical powers of the lens to be implanted (knowing surgically induced astigmatism), and finish with the marking of the axis and precise alignment of the lens, in addition to a surgical technique that must be meticulous.

This chapter focuses on the process of calculating these lenses, focusing on the sources of error of traditional calculators and the new strategies that have been developed to overcome them.

TORIC LENS CALCULATORS – SOURCES OF ERROR

In 1992, Shimizu et al designed the first toric IOL. After the introduction of these lenses in the market and the expansion of their use, several sources of error were identified in the calculators made available online by the manufacturers. Some of them have, however, been updated in order to correct these errors, although many other known errors remain uncorrected.

First, it is known that for each cylindrical power in the plane of the IOL, a variable power of astigmatism is corrected in the plane of the cornea. This variability depends on the distance between the cornea and the IOL7. The classical calculators assumed that the relationship between the cylindrical power in the corneal plane and the IOL was fixed. In the original Acrysof toric calculator (Alcon Laboratories Inc., Fort Worth, TX, USA) a ratio of 1.46 was considered.

Although this factor may be of little relevance to eyes with average axial lengths, their use results in hypocorrections in long eyes and hypercorrections in short eyes (for example, for an eye with an axial length of 20.0 mm, the actual ratio is 1.29, being 1.86 for an eye of 30.0 mm)10,11. Several strategies have been suggested to overcome this limitation, such as the inclusion of anterior chamber depth and pachymetry in the calculation of IOL or the use of Fam and Lim’s meridional analysis12,13.

In addition to the first limitation, although the scientific literature is scarce on the subject, the cylindrical power of IOL in the plane of the cornea is not independent of its spherical power, due to the different vergence of the rays. Considering the same effective position of the lens, take as an example an Alcon Acrysof SN60T3 lens (1.50 D of cylindrical power in the plane of the IOL and 1.03 D in the corneal plane, according to the manufacturer), the actual cylindrical power in the corneal plane is 1.32 D for a lens of +17.00 D and 1.22 D for a lens of +28.00 D.

In the case of an SN60T9, the cylindrical power in the plane of the cornea is 5.28 D for an IOL of +17.00 D and 4.88 D for an IOL of +28.00 D (not 4.11 D as suggested by the manufacturer). Failure to consider spherical power induces errors which, in the case of an SN60T9, may be greater than 1.00 D12.

Traditionally, keratotomes and topographers only evaluate the anterior face of the cornea, assuming a fixed relationship between the curvature of the anterior and posterior faces, and a keratometric index of 1.3375 is generally used to convert measurements of the anterior face in total corneal power and astigmatism. It is known that this index is not correct since it overestimates corneal power by about 0.56 D14.

With the appearance of tomography instruments, such as the Scheimpflug chambers, which allow direct evaluation of both faces of the cornea, the importance of the posterior face of the cornea, which until then was thought to induce little astigmatism and could be ignored in the calculation of toric lenses, was recognized.

In 85% of cases, its more curved meridian vertically aligned generates positive power in the horizontal meridian, so selecting toric IOLs based only on measurements of the anterior face leads to hypercorrections in eyes with astigmatism in favor of the rule (WTR) and hypocorrections in eyes with posterior astigmatism when this cannot be directly measured (Figure 2).

Since then, several nomograms have been developed to consider the posterior corneal face in the calculation of toric IOLs, not measuring it directly as the calculators described above that exceed the limitations, especially because it is now well known that failure to consider the posterior face of the cornea is the most important error factor17.

NEW CALCULATORS AND COMPARISON OF THEIR RESULTS

In order to compare the new calculation methodologies to overcome all or some of the sources of error mentioned above (Table 1), a study was recently carried out on 86 eyes with Acrysof toric lenses, in which the error of prediction of the residual astigmatism of each of these methods was calculated18.

In this study it was concluded that the methods with lower average error and lower dispersion of prediction of errors were Barrett's calculator and the Abulafia-Koch formula (Figure 3a and b), which reduced the residual astigmatism prediction error of the original Alcon calculator by more than 50%. It was also verified that these two nomograms presented errors inferior to the use of real measurements from a Scheimpflug’s chamber (Figure 4).

Subsequently, the results of measures estimated with actual measures of the posterior surface of the cornea were compared (Table 2)¹⁹.

Once again, the centroids of the error were found to be lower in the estimation methods than in those using actual measurements of the corneal posterior face (Table 3).

It was observed not only in the total eyes studied, but also when subdivided by type of astigmatism (WTR/ATR), although the differences between the estimation methods and actual measurements were higher in eyes with astigmatism WTR, which indicates that Pentacam may underestimate the power of the posterior surface of the cornea in eyes with WTR astigmatism.

Other studies confirm that Pentacam and other Scheimpflug-based tomographs may underestimate vertical posterior astigmatism and overestimate it in the horizontal meridian20,21.

In addition, the lower reproducibility of the Scheimpflug chambers is known in the assessment of the posterior face of the cornea relative to what is possible for the anterior surface22.

The estimation methods also allowed obtaining a higher percentage of eyes within lower prediction errors (Figure 5).