Ultrasound Biometry

AL has traditionally been measured using ultrasound biometry. With the applanation technique, the ultrasound probe is placed in direct contact with the cornea, and corneal compression typically causes the AL to be falsely shortened. Applanation biometry has given way to noncontact methods. Although the immersion technique has been shown to be more reproducible than the applanation technique, both require mindfulness of the properties of ultrasound. In eyes with high to extreme axial myopia, the presence of a posterior staphyloma should be considered. Erroneously long AL readings may occur in eyes with staphylomata. An immersion A/B-scan approach for AL measurement has been described in the setting of posterior staphyloma.

With A-scan biometry, errors in AL measurement account for 54% of IOL power error when two-variable formulas are used.

Optical Biometry

Optical biometry has been shown to be significantly more accurate and reproducible and is rapidly becoming the most prevalent methodology for the measurement of AL. The most commonly used optical biometers are IOLMaster (Carl Zeiss Meditec, Jena, Germany) and Lenstar (Haag-Streit, Koeniz, Switzerland).

• •
  

  IOLMaster: The IOLMaster 500 was introduced in 2000 as the first optical biometer. Based on partial coherence interferometry technology, it uses a 780-nm laser diode to measure AL. The device also provides measurements of keratometry, anterior chamber depth (ACD), and white-to-white (WTW) distance.

  
  
  
  
  • 
    

    The newer version of this device (IOLMaster 700) uses an optical configuration that allows telecentric and thus distance-independent keratometry measurement. It uses swept-source optical coherence tomography (OCT) to is measure axial length, central corneal thickness (CCT), and lens thickness (LT). It displays a full-length OCT image, showing anatomical details of a longitudinal cuts through the entire eye ( Fig. 5.5.1 ).

    
    

    
    

    
    

    

    
    

    
    

    Display From the Swept-Source Optical Coherence Tomography Biometer IOLMaster 700.

    
    

  
  
  
  
• •
  

  Lenstar: Based on optical low-coherence reflectometry technology, the Lenstar uses an 820-nm laser diode to measure AL, ACD, CCT, and LT. It calculates keratometry from an array of 32 light reflections projected off the anterior corneal surface.

  
  
• •
  

  Argos (Movu Inc, Komaki, Japan) and OA-2000 (Tomey, Nagoya,Japan): These are two other new swept-source biometers that recently been introduced.

Studies have shown that the repeatability of the IOLMaster and Lenstar for all biometric parameter measurements is excellent and that agreement between these devices is good. The differences in AL, ACD, and LT between these devices were not shown to produce a statistically significant difference in IOL power calculation.

Although measurements with optical biometry are mostly operator independent, careful alignment during the scan and inspection of the measurement quality are still necessary for optimal refractive outcomes ( Fig. 5.5.2 ).

Poorly reflected LED images from the anterior corneal surface using IOLMaster 700 (top); poorly reflected LED images from the anterior corneal surface using Lenstar with average keratometry and corneal astigmatism of 43.24 D and 2.05 D@174° (middle), and repeated measurements with average keratometry and corneal astigmatism of 44.36 D and 0.25 D@167° (bottom).

The primary limitation of optical biometry is its inability to measure through dense cataracts and other media opacities that obscure the macula. It was reported that with use of an earlier generation of IOLMaster, approximately 10% of eyes could not be accurately measured due to such opacities or fixation difficulties. The IOLMaster 700 showed better penetration in dense posterior subcapsular cataracts, measuring AL successfully in 96% of cases.

Vergence Formulas

The majority of IOL formulas are vergence-based formulas. Based on the number of variables they use to calculate ELP, these formulas can be categorized into the following groups:

• •
  

  Two-variable formulas: These include the Holladay 1, Hoffer Q, and SRK/T, and they use AL and keratometry to calculate the distance from the principal plane of the cornea to the thin lens equivalent of the IOL (i.e., ELP). Thus, a short eye or an eye with a flatter cornea will have a shallower anterior chamber. However, Holladay has shown that exceptions to these assumptions exist.

  
  
• •
  

  Three-variable formula: The Haigis formula uses AL, keratometry, and ACD.

  
  
• •
  

  Five-variable formula: The Barrett Universal II formula uses AL, keratometry, ACD, LT, and WTW.

  
  
• •
  

  Seven-variable formula: The Holladay 2 formula uses preoperative refraction, age, AL, keratometry, ACD, LT, and WTW.

Ray Tracing Formulas

• •
  

  PhacoOptics: With PhacoOptics, IOL power is calculated based on exact ray tracing (Snell’s law of refraction). It incorporates the latest generation ACD prediction algorithms based on the complex relationship between the preoperative ocular dimensions (ACD and LT) and the postoperative position of the IOL (postoperative ACD). Measurements of the anterior and posterior corneal curvatures as along with conic coefficients (Q-values) obtained by modern anterior segment imaging systems can be used directly by the program.

  
  
• •
  

  Okulix: Okulix is a program package that calculates single rays using Snell’s law. AL can be entered either manually or by a computer link to the measuring device. As an alternative to entering corneal radii by hand, they also can be taken from a two-dimensional corneal topographic map.

Artificial Intelligence Formulas

• •
  

  Radial basis function (RBF): The Hill-RBF calculator is an advanced, self-validating method for IOL power selection employing pattern recognition and a sophisticated form of data interpolation. Based on artificial intelligence, this methodology is entirely data driven. Pattern recognition for selecting an IOL power is achieved through the process of adaptive learning, the ability to learn tasks based solely on data. Unlike static theoretical formulas, this approach will be an ongoing project and continuously updated as a “big data” exercise. The greater the number of surgical outcomes that are fit to the RBF model, the greater the overall depth of accuracy.

  
  
• •
  

  Neural network: Clarke developed a computational method based on neural network in which the software is trained to predict IOL powers using large amount of clinical data from one surgeon with one IOL. Clinical data include preoperative AL, keratometry, ACD, and LT.

Combination Formulas

• •
  

  Super formula: The Ladas super formula amalgamates outcomes from the above-mentioned two-variable and three-variable vergence formulas and has a small component of artificial intelligence.

All of these formulas have some element of regression, as they include constants that are derived from prior patient outcomes. Note that two critical data points are not measured: posterior corneal curvature (although this is slowly being integrated) and ELP. Better predictive formulas for estimating ELP will likely require more sophisticated measurements, possibly to include lens diameter, lens volume, and certain angle and iris features.

IOL Power Calculation in Short Eyes

In short eyes, the importance of accurate ELP prediction is magnified due to the high power of the IOL and the relatively short distance from the IOL to the retina. Olsen showed that 0.25 mm error in postoperative ACD corresponds to a 0.1 diopter (D) error in long eyes with 30 mm AL and a 0.5 D error in short eyes with 20 mm AL.

In IOL calculation formulas that do not use ACD in the ELP calculation, it is assumed that short eyes have a proportionally shallower anterior chamber. However, this assumption breaks down in the many short eyes that in fact have normal anterior chamber anatomy with normal ACD.

Several studies in the literature have compared different IOL power formulas and their accuracy in short eyes. Although it is believed that newer IOL power prediction formulas perform best in short eyes, uncertainty still exists in the literature on whether any of the available formulas perform better than the others.

In general, our approach for short eyes is to use the Holladay 1, Holladay 2, Hill-RBF, Olsen, and Barrett, favoring the Olsen when there is disagreement. We also try to operate on the nondominant eye first so that we can use its refractive outcome to adjust the IOL power for the second eye, generally changing the calculated IOL power by one-half of the prediction error in the first eye.

IOL Power Calculation in Long Eyes

In long eyes, IOL power formulas tend to select IOLs of insufficient power, leaving patients with postoperative hyperopia. Inaccurate measurement of preoperative AL has been reported to be the main reason for postoperative refractive error in axial high myopia. The incidence of posterior staphyloma increases with increasing AL. Ultrasonic biometric methods can produce errors in the presence of a posterior staphyloma by giving falsely long AL.

Theoretically, optical biometry permits more accurate measurements when a posterior staphyloma is present. However, in a study investigating the accuracy of SRK/T formula in eyes with negative or zero-powered IOLs, MacLaren and colleagues reported consistent hyperopic errors across all three methods of biometry (A-scan, B-scan, and optical). This indicates that eliminating or minimizing the adverse impact of posterior staphylomata on IOL calculations does not prevent hyperopic surprises in long eyes.

We proposed a method for optimizing AL in long eyes (Wang–Koch adjustment). Our results showed that this method significantly improved the accuracy of IOL power calculation in eyes with IOL powers ≤5.00 D and significantly reduced the percentage of eyes that would be left hyperopic. We recommend using the optimized AL in eyes with AL >25.2 mm. Based on the formula, the optimized AL is calculated from the measured optical or ultrasonic AL using the following equation:

Then the optimized AL is entered into the IOLMaster or Lenstar and the calculation is performed again. We recommend selecting the IOL power that predicts a minus prediction error close to zero (−0.1 to −0.2 D), since slight myopic results may occur with this approach of optimizing AL.

A recent advance has been the development of the Barrett Universal II formula, which has been refined to improve outcomes in long eyes. In a study of eyes with AL ≥26.0 mm, Abulafia et al. reported that for IOL powers <6.00 D, best results occurred with the Holladay 1 with Wang–Koch adjustment, Haigis with Wang–Koch adjustment, and Barrett Universal II formulas.

Due to the low IOL powers required in long eyes, accuracy of ELP estimation is not as important as in normal and short eyes. By refining the AL value used in current formulas, excellent outcomes can be anticipated as shown in these studies.

IOL Power Calculation in Eyes With Previous Corneal Refractive Surgery

Cataract surgeons are facing challenges in IOL power calculation in eyes that have undergone excimer laser photorefractive keratectomy (PRK), laser in situ keratomileusis (LASIK), or radial keratotomy (RK).

Methods to Improve the Accuracy of IOL Power Calculations in Postrefractive Eyes

Many approaches have been proposed to improve the accuracy of IOL power calculation in eyes following LASIK/PRK/RK. These can be categorized into three groups depending on the use of historical data acquired before refractive surgery was performed.

Methods Relying on Prior Clinical Data

Methods in this category use completely historical data. Clinical studies have shown that they are less accurate than the formulas in other categories described in the following sections. The concern with these methods is their sensitivity to errors in the historically obtained data. A 1.00 D error in either the keratometric or refractive values translates to nearly a 1.00 D postoperative refractive error.

Methods Using a Combination of the Surgically Induced Refractive Change (ΔMR) and Current Corneal Power Values

These methods modify either corneal power measurements at the time the patient presents for cataract surgery or calculated IOL power based on ΔMR.

These methods multiply ΔMR by a fraction of between 0.15 and 0.33, depending on the formula. This translates to an error of 0.15–0.33 D for each 1.00 D of error in ΔMR, reducing potential errors caused by having inaccurate historical data. Studies have shown that some of these methods have consistently been among the more accurate approaches.

Methods Requiring No Historical Data

Several methods requiring no historical data have been proposed. Surgeons use these approaches most often. The formulas fall into two categories:

• •
  

  Formulas that adjust measured corneal power from the anterior corneal surface based on either regression analysis or assumed posterior corneal power, such as the Wang–Koch–Maloney, Shammas, Haigis-L, Potvin-Hill Pentacam, and Barrett True K No History formulas.

  
  
• •
  

  Formulas based on corneal power measurements from both anterior and posterior corneal surfaces. Using the RTVue (Optovue Inc, Fremont, CA), Tang and colleagues developed an OCT-based IOL calculation formula using the anterior and posterior corneal powers and the central corneal thickness.

Methods requiring no historical data have been shown to perform as well as those using a combination of ΔMR and current corneal power values. Promising results for the OCT-based IOL formula and Barrett True K formula have been reported. However, the percentage of eyes within 0.5 D of target refraction was under 70% for all formulas, the highest value being 68.3% for the OCT formula. Obviously, more studies are needed to improve the accuracy of corneal power measurements and to develop new IOL power calculation formulas in these eyes.

Web-Based Post-Refractive IOL Calculator

To simplify the complicated and time-consuming calculations discussed above, we developed a web-based post-refractive IOL power calculator in 2007 ( Fig. 5.5.3 ) ( http://www.ascrs.org/ ). This calculator can be used for eyes with previous myopic LASIK/PRK, hyperopic LASIK/PRK, or RK. We have performed major updates to the online calculator in the past and will continue to update it. During the past year, the number of visits to this calculator was over 120,000.

Postrefractive IOL Calculator at American Society of Cataract and Refractive Surgery ( http://www.ascrs.org/ ).

Radial Keratotomy

IOL power calculation in RK eyes is even more difficult due to the greater irregularity of the anterior corneal curvature and the posterior corneal curvature changes. Furthermore, it has been reported that 20%–50% of RK eyes have a gradual hyperopic shift. We recommend using topographically derived average corneal power over the central 2–4 mm zone. Compensation for potential error in ELP is still required, with use of double-K version of IOL formulas if the corneal power is used for ELP prediction.

We find relatively lower accuracy in post-RK eyes compared with eyes following LASIK/PRK. In 95 post-RK eyes, we evaluated the accuracy of newer IOL formulas (OCT and Barrett True K), and the percentage of eyes within 0.5 D of target was less than 50% for all formulas. Further improvements in the accuracy of IOL power calculation in RK eyes are desirable and await robust techniques to measure anterior and posterior corneal curvature.

IOL Power Calculation in Keratoconic Eyes

IOL calculations are more difficult in eyes with keratoconus, presumably due to the irregularity of the cornea and the change in ratio of anterior to posterior corneal curvatures. Using the Holladay 1, Barrett, and Olsen formulas in 21 keratoconic eyes, we found that the mean refractive prediction errors were +1.25 D, +1.15 D, and +1.36 D, respectively, and the magnitude of the hyperopic prediction error with Holladay 1 formula increased with increasing corneal power. Improved corneal imaging technology for accurate anterior and posterior corneal power measurements is needed.

IOL Power Calculation in Eyes Following Keratoplasty

Eyes following penetrating keratoplasty may have high amounts of irregular anterior corneal astigmatism and uncertain posterior corneal values. With Descemet’s stripping automated endothelial keratoplasty, corneal power changes are less pronounced with minimal astigmatic change but with a hyperopic shift of around 0.70–1.5 D. Descemet membrane endothelial keratoplasty induces even more modest refractive changes, with a reported hyperopic refractive shift of 0.24–0.50 D and minimal change in refractive astigmatism. Although the complexity of IOL calculation has diminished as procedures have advanced, it remains challenging to perform IOL power calculation in these eyes.