Corneal ectasia is a multifactorial condition in which the cornea progressively steepens and thins. The most common corneal ectasia is keratoconus, but it can also occur after excimer laser surgeries such as LASIK and PRK, radial keratotomy, and small incision lenticule extraction (SMILE).
Progression of corneal ectasia induces myopia and irregular astigmatism. Because corneal power varies greatly within a small area, IOL power calculation becomes significantly more difficult than in normal eyes 1). The main reasons are the following two points.
When standard IOL formulas are used, keratoconic eyes tend to have hyperopic surprise postoperatively. Therefore, the use of keratoconus-specific formulas is recommended 1).
Cataract surgery can be performed even in the presence of corneal ectasia. However, because the accuracy of IOL power calculation decreases, specialized formulas and detailed examination using corneal tomography are necessary 1). Preoperative confirmation of corneal stability is a prerequisite.
Visual symptoms associated with corneal ectasia include the following:
Clinical findings of corneal ectasia are evaluated by corneal topography and tomography.
Since visual function cannot be fully assessed by corrected visual acuity alone, comprehensive evaluation including corneal topography and tomography is necessary.
The difficulty in calculating intraocular lens power is mainly attributed to the following three factors 1).
Third- and fourth-generation intraocular lens formulas, except for the Haigis formula, use corneal refractive power to calculate the predicted postoperative anterior chamber depth. In eyes with corneal flattening (after myopic LASIK) or steepening (ectasia), underestimation or overestimation of the effective lens position occurs 1).
Ectatic eyes tend to have long axial lengths and deep anterior chambers. Ultrasound biometry is inaccurate; optical biometry, which allows measurement along the visual axis, is recommended.
Standard formulas assume a normal corneal shape. In eyes with ectasia, the curvature ratio of the anterior and posterior corneal surfaces changes, making the K value inaccurate. Furthermore, the algorithm that predicts the effective lens position from the K value also introduces errors, leading to postoperative hyperopia 1).
In patients with ectasia being considered for cataract surgery, manifest refraction is the first step. This allows the surgeon and patient to share an estimate of the expected postoperative outcome.
Hard contact lens over-refraction is useful for differentiating whether the cause of reduced vision is corneal or cataractous. If vision improves with a hard CL, the cause is corneal; if not, it is cataractous.
The accuracy of corneal refractive power measurement determines the success of intraocular lens calculation. The main measurement methods are shown below.
| Measurement Method | Measured Surface | Notes |
|---|---|---|
| Manual/Automated Keratometer | Anterior surface only | Central 3 mm; posterior surface is estimated |
| Corneal Topography | Anterior surface only | Slope → calculates elevation differences |
| Corneal tomography | Anterior + posterior | Gold standard |
In keratoconus eyes, anterior corneal astigmatism is often with-the-rule, and posterior corneal astigmatism is often against-the-rule 1). It is recommended to evaluate anterior, posterior, and total corneal astigmatism 1). A cutoff of 1.8 D for anterior corneal astigmatism is useful for differentiating keratoconus from normal eyes, with reported sensitivity and specificity of 90.2% 1).
Optical biometry is recommended for axial length measurement. Representative devices are as follows.
For Pentacam, the “True net power map” and “equivalent K value” are useful. The True net power map is calculated by assigning individual refractive indices to the anterior and posterior corneal surfaces. The equivalent K value allows checking K values in any zone from 1 to 7 mm, helping to select the K value used for intraocular lens calculation.
As a prerequisite for intraocular lens power calculation, confirmation of corneal stability is essential. Performing calculations on a progressing cornea will yield poor results.
Typically, keratoconus does not progress after age 50. However, pellucid marginal degeneration may continue to progress, so caution is required 1).
Dedicated Formulas (Recommended)
Barrett True-K keratoconus formula: Can use measured posterior corneal curvature. Shows good accuracy even in severe cases 1).
Kane keratoconus formula: A formula utilizing artificial intelligence (AI). Provides more accurate refractive target prediction than traditional formulas 1).
EVO 2.0 (using TK): Improved accuracy in moderate keratoconus 1).
Traditional Formulas
SRK/T: Shows the best performance among traditional formulas 1). However, it is inferior to dedicated formulas.
Others: Hoffer Q, Holladay 1 & 2, Haigis, etc., have a strong tendency toward hyperopic shift. Traditional formulas other than SRK/T are recommended to be avoided in keratoconic eyes 1).
In a systematic review, the mean absolute errors of each formula were reported as: Barrett Universal II 0.314D (82.1%), Haigis 0.346D (76.1%), Holladay 2 0.351D (69.1%), SRK/T 0.389D (71.3%), Hoffer Q 0.409D (63.3%), and Holladay 1 0.409D (62.0%) 1). The percentages in parentheses indicate the proportion within ±1.0D of the target refraction.
For keratoconus eyes with K value ≤55 D, a slightly myopic target (-0.5 D to -1.5 D) is recommended 1). This accounts for the risk of postoperative hyperopic shift. Similarly, a myopic target is recommended for eyes after radial keratotomy 1).
In Japanese clinical practice, the Double-K method, the OKLIKUS ray-tracing software from the anterior segment OCT (Tomey), the Calmellin-Calossi formula in IOL-Station (Nidek), and the Haigis-L formula are considered useful. The American Society of Cataract and Refractive Surgery (ASCRS) website provides a free IOL calculator that allows simultaneous comparison of multiple formulas. The Barrett True K formula, updated in 2015, also supports IOL power calculation after hyperopic LASIK and radial keratotomy.
The Barrett True-K keratoconus formula and the Kane keratoconus formula are recommended 1). If traditional formulas must be used, SRK/T shows relatively good results. It is advisable to calculate with multiple formulas and compare the results.
It is useful for correcting regular astigmatism in cases with stable corneas 1). However, complete correction of irregular astigmatism is difficult, and toric contact lenses cannot be used postoperatively. It is not indicated for patients who plan to wear hard contact lenses 1).
This section explains in detail the mechanisms of intraocular lens power calculation errors.
The three basic measurements required for intraocular lens power calculation are axial length, corneal refractive power (K value), and effective lens position 1). In eyes with ectasia, errors can occur in all of these.
Total corneal refractive power is determined by the sum of the refractive powers of the anterior and posterior corneal surfaces. Standard keratometry and topography measure only the anterior corneal surface and use a corneal refractive index of 1.3375 to estimate the posterior curvature. In normal eyes, this approximation is valid because the ratio of anterior to posterior curvature is constant, but in ectatic eyes, the ratio changes, leading to errors 1).
In keratoconic eyes, anterior corneal astigmatism is predominantly against-the-rule, while posterior corneal astigmatism is predominantly with-the-rule 1). Automated keratometers have been reported to have a bias of overcorrection for with-the-rule astigmatism and undercorrection for against-the-rule astigmatism.
IOL prediction accuracy is affected by the severity of keratoconus 1). In mild to moderate cases, newer formulas (e.g., EVO 2.0 TK) improve accuracy, but in advanced cases, performance is relatively low with any traditional formula 1). In advanced cases, the use of Barrett True-K or Kane keratoconus formulas is particularly recommended 1).
Severity classifications include the Amsler-Krumeich classification, Alio-Shabayek classification, and Belin ABCD grading system 1).
Development of intraocular lens calculation formulas utilizing artificial intelligence and machine learning, such as the Hill-RBF formula and Kane formula, is progressing. These analyze large datasets and build predictive models adapted to the anatomical and refractive characteristics of individual eyes. Improved accuracy has been reported, especially in eyes with abnormal axial length.
Although the superiority of formulas dedicated to keratoconus has been demonstrated, subgroup analysis by severity is limited by the number of cases1). Definitive conclusions require large-scale studies1). Advances in devices and IOL technology are improving the safety and predictability of calculations.
