Refractive outcomes of intraocular lens power calculation using different corneal power measurements with a new optical biometer

Refractive outcomes of intraocular lens power calculation using different corneal power measurements with a new optical biometer

1 ARTICLE Refractive outcomes of intraocular lens power calculation using different corneal power measurements with a new optical biometer Giacomo S...

463KB Sizes 0 Downloads 68 Views

1

ARTICLE

Refractive outcomes of intraocular lens power calculation using different corneal power measurements with a new optical biometer Giacomo Savini, MD, Kazuno Negishi, MD, Kenneth J. Hoffer, MD, FACS, Domenico Schiano Lomoriello, MD

Purpose: To evaluate the results of intraocular lens (IOL) power calculation using different corneal power measurements provided by an optical biometer combined with a dual Scheimpflug analyzer and a Placido disk topographer (Galilei G6). Setting: G.B. Bietti Foundation, Rome, Italy, and Keio University Hospital, Tokyo, Japan.

Design: Evaluation of diagnostic technology. Methods: Consecutive patients having cataract surgery were enrolled. The IOL power was calculated with the Barrett Universal II, Haigis, Hoffer Q, Holladay 1, and SRK/T formulas. Different options were used to calculate the corneal power: simulated keratometry (K) based on anterior corneal surface measurements only and total corneal power (TCP) based on ray tracing through both corneal surfaces. Three TCP measurements (TCP1, TCP2, and TCP-IOL) were evaluated.

S

ince optical biometry was introduced in 1999 with the IOLMaster (Carl Zeiss Meditec AG), many manufacturers have developed newer optical biometers to calculate intraocular lens (IOL) power. The first alternative was the LenstarLS-900 (Haag-Streit AG) in 2009. It was followed by the Aladdin (Topcon Europe Medical B.V./ Visia Imaging S.r.l.), the AL-Scan (Nidek Co. Ltd.), the Galilei G6 (Ziemer Ophthalmology GmbH), the OA2000 (Tomey Corp.), the Argos (Movu, Inc.), and the Pentacam AXL (Oculus Optikger€ate GmbH). Several papers have investigated the accuracy of IOL power calculation using the IOLMaster (in its various versions, including the IOLMaster 500 and 700)1–9 and Lenstar LS900,2,8,10,11 whereas only a few have reported the results with the other devices.12–16 Recently, Ventura et al.13 provided us with the results of the Galilei G6, the first optical

Results: The study analyzed 118 eyes. The mean values of simulated K (43.74 diopters [D] G 1.40 [SD]), TCP1 (43.13 G 1.35 D), TCP2 (41.87 G 1.30 D), and TCP-IOL (42.62 G 1.35 D) were significantly different (P < .0001). The best results were obtained using simulated K: the median absolute error ranged between 0.22 D and 0.29 D and the percentage of eyes with a prediction error of G0.50 D or less, between 76.2% and 84.7%, depending on the formula. After constant optimization, the results using any TCP value and simulated K were similar with no statistically significant differences.

Conclusions: Biometric measurements provided by the Scheimpflug–Placido optical biometer can be used to accurately calculate the IOL power. Simulated K and TCP led to similar outcomes after constant optimization. J Cataract Refract Surg 2018; -:-–- Q 2018 ASCRS and ESCRS

biometer combined with a dual Scheimpflug camera. The outcomes were good, but they were obtained from a small sample (only 37 eyes), with 1 formula (the Haigis) using the optimized constants from the IOLMaster and were based on only 1 keratometry (K) method (simulated K). However, the Galilei G6 (referred to hereafter as the Scheimpflug–Placido optical biometer) offers different modalities to calculate the IOL power, namely total corneal power 1 (TCP1), total corneal power 2 (TCP2), and total corneal power IOL (TCP-IOL), all derived from corneal ray tracing; their accuracy for IOL power calculation has not been investigated and reported. This study was designed with 4 purposes: (1) to report the results of IOL power calculation by means of the Scheimpflug–Placido optical biometer on a larger sample than previously reported, (2) to assess which corneal power

Submitted: November 27, 2017 | Final revision submitted: March 16, 2018 | Accepted: March 20, 2018 From the G.B. Bietti Foundation IRCCS (Savini, Schiano Lomoriello), Rome, Italy; the Department of Ophthalmology (Negishi), Keio University School of Medicine, Hospital, Tokyo, Japan; Stein Eye Institute (Hoffer), University of California, Los Angeles, and St. Mary’s Eye Center (Hoffer), Santa Monica, California, USA. The contribution of G.B. Bietti Foundation IRCCS was supported by the Italian Ministry of Health and Fondazione Roma, Rome, Italy. Graham D. Barrett, MD, optimized the lens factor of the Barrett Universal II formula. Corresponding author: Giacomo Savini, MD, Fondazione G.B. Bietti Foundation IRCCS, Via Livenza 3, Rome, Italy. Email: [email protected]. Q 2018 ASCRS and ESCRS Published by Elsevier Inc.

0886-3350/$ - see frontmatter https://doi.org/10.1016/j.jcrs.2018.03.027

2

IOL POWER CALCULATION WITH A NEW BIOMETER

modality enables the most accurate results for IOL power calculation, (3) to report the optimized constants for the new biometer and compare them with those reported for the IOLMaster in the User Group for Laser Interference Biometry (ULIB) website,A and (4) to verify whether the outcomes are different between white and Asian eyes. PATIENTS AND METHODS This was a prospective interventional multicenter study. Consecutive patients having cataract surgery and implanted with nontoric single focal IOLs were enrolled between March 2016 and October 2017 at 2 separate centers: G.B. Bietti Foundation IRCCS (Rome, Italy) and Keio University Hospital (Tokyo, Japan). Before being included in the study, all patients were informed of its purpose and gave their written consent. The study protocol was approved by the Ethics Committees of both institutions. Exclusion criteria were previous corneal or intraocular surgery, keratoconus and any other corneal disease, contact lens usage during the previous month, and postoperative corrected distance visual acuity (CDVA) lower than 0.8 (20/25) for any reason. Patients were also excluded when optical biometry measurements were not possible because of lens opacities (in this event, lens opacities were graded according to the Lens Opacities Classification System III [LOCS III]).17 Phacoemulsification was performed by 2 surgeons (G.S., K.N.) through a 2.75 mm temporal incision under topical anesthesia. All patients received the same IOL model (Acrysof SN60WF, Alcon Laboratories, Inc.), so that formula constants optimization could be carried out as recently suggested by Hoffer et al.18 Before surgery, all patients had optical biometry with the Scheimpflug–Placido optical biometer. This measures axial length (AL) and lens thickness by means of an 880 nm A-scan interferometer. Based on a proprietary algorithm, it derives anterior segment measurements (anterior and posterior corneal curvature, corneal thickness, and diameter) from the images obtained by 20 Placido rings and 2 oppositely rotating Scheimpflug cameras. The axial distance from the corneal epithelium to the anterior surface of the lens is measured by means of the Scheimpflug cameras. Corneal power values are calculated based on the anterior (and posterior) corneal curvatures in the 1.0 to 4.0 mm central zone. The following corneal powers calculated by the optical biometer were used for IOL power calculation (the definitions of TCP are those provided in the Operator Manual of the Scheimpflug– Placido optical biometer):  Simulated K: This is calculated by converting the measured radius of the corneal surface into diopters (D) using the standard 1.3375 keratometric refractive index.  Total corneal power 1: This is calculated using the corneal index of refraction (n Z 1.376). For the purpose of determining the focal length, the defined reference plane is the anterior corneal surface.  Total corneal power 2: This is calculated using the aqueous index of refraction (n Z 1.336). For the purpose of determining the focal length, the defined reference plane is also the anterior corneal surface.  Total corneal power–IOL is calculated using the aqueous index of refraction (n Z 1.336). For the purpose of determining the focal length, the defined reference plane is the posterior corneal surface. The IOL power was calculated according to the Haigis,19 Hoffer Q, Holladay 1,22 and SRK/T formulas.23 Moreover, the Barrett Universal IIB was used to calculate the IOL power with simulated K (but not with TCP because this formula is designed to work only with corneal powers obtained by the keratometric refractive index). A final evaluation was performed by assessing the postoperative subjective refractive outcomes 1 month after surgery, which 20,21

Volume - Issue - - 2018

is when refractive stability can be expected with small-incision clear cornea surgery and this type of IOL.24 Postoperative subjective refraction was measured at 4 m and then adjusted to infinity by subtracting 0.25 D, as recommended by Simpson and Charman.25 To calculate the prediction error in refraction, the measured manifest refractive spherical equivalent was subtracted from the predicted refraction (based on the IOL power actually implanted) according to each formula. The mean arithmetic error, the median absolute error, and the mean absolute error (MAE) were calculated, as well as the rate of eyes with a prediction error of G0.50 D or less.18,26,27 Predictions made using each formula were optimized in retrospect by adjusting the respective constants to give an arithmetic prediction error of zero in the average case, according to the method described by Hoffer20,21,27 and Olsen.4 Optimization of the Barrett Universal II formula was carried out personally by Graham D. Barrett, MD, who provided the lens factor for each sample. Barrett’s formula calculations were carried out through the website.B As a result of constant optimization, it was possible to evaluate the statistical error as representing the optimum prediction error rather than offset errors related to incorrect lens constants or systematic errors in the measuring environment. Statistical Analysis An unpaired t test was used to compare the mean values of the measured parameters between the 2 samples and repeated measures analysis of variance (ANOVA) was used to compare the mean values of the 4 corneal power modalities within each sample. Friedman’s test (nonparametric ANOVA) with Dunn’s post-test was performed to compare the mean values of corneal power, median absolute errors, and MAEs in refractive outcome. The chisquare test was used to compare the percentage of eyes within G0.50 D or less than the predicted refraction. A P value less than 0.05 was considered statistically significant. For patients who had bilateral surgery, only the first eye operated on was considered for statistical analysis. All statistical analyses were carried out using Instat software (version 3.1, Graphpad Software, Inc.) and Medcalc (version 12.3.0, Medcalc Software, Inc.). Based on power and sample size calculations performed using Medcalc, it was estimated that a sample size of 99 eyes would be necessary to detect a difference of 0.05 D between simulated K and TCP with a power of 95% at a significance level of 5%, given a within-subject standard deviation for simulated K and TCP equal to 0.12 D and 0.13 D, respectively.28

RESULTS In the Italian group, 71 eyes of 71 patients were consecutively enrolled; the AL of 3 of the patients could not be measured because of lens opacities. Therefore, the analysis was carried out on 68 eyes of 68 patients (35 men; mean age: 73.6 years G 8.9 [SD]). All 3 patients had a nuclear cataract (N) with posterior subcapsular (P) opacities (stages NIII and PIII according to the slitlamp-based LOCS III). In the Japanese group, 50 eyes of 50 patients (26 men; mean age: 71.2 G 7.8 years) were enrolled, with no cases excluded. Thus, a total of 118 eyes of 118 patients were analyzed. Table 1 shows the measurements in the Italian group and the Japanese group. The unpaired t test did not disclose any statistically significant difference between the 2 groups for any parameter. In the whole sample as well as in both subsamples, repeated measures ANOVA showed that the difference between the 4 corneal power values was statistically significant (P ! .0001), with simulated K and TCP 2 providing, respectively, the highest and lowest mean values.

3

IOL POWER CALCULATION WITH A NEW BIOMETER

Table 1. Measured parameters in the analyzed samples. Total Cohort (N Z 118) Parameter AL (mm) ACD (mm) Lens thickness (mm) Sim-K (D) TCP1 (D)† TCP2 (D)z TCP-IOL (D)x

Mean ± SD 24.19 G 1.31 3.24 G 0.39 4.46 G 0.40 43.74 G 1.40 43.13 G 1.35 41.87 G 1.30 42.62 G 1.35

Range 21.63, 28.65 2.33, 4.20 3.33, 5.22 40.58, 47.63 39.82, 46.97 38.66, 45.61 39.31, 46.51

Italian Group (n Z 68) Mean ± SD 24.07 G 1.25 3.26 G 0.42 4.50 G 0.47 43.63 G 1.27 43.08 G 1.21 41.84 G 1.18 42.58 G 1.22

Range 22.15, 28.34 2.33, 4.20 3.33, 5.22 40.58, 46.19 39.82, 45.53 38.66, 44.21 39.31, 45.03

Japanese Group (n Z 50) Mean ± SD 24.35 G 1.41 3.22 G 0.35 4.42 G 0.32 43.88 G 1.57 43.18 G 1.53 41.92 G 1.46 42.67 G 1.53

Range 21.63, 28.65 2.49, 4.08 3.61, 5.21 40.87, 47.63 40.18, 46.97 39.01, 45.61 39.69, 46.51

P Value* .2530 .5716 .3446 .3304 .6843 .7371 .7012

ACD Z anterior chamber depth; AL Z axial length; IOL Z intraocular lens; Sim-K Z simulated keratometry; TCP Z total corneal power *Unpaired t test † Calculated using the corneal index of refraction (n Z 1.376); for the purpose of determining the focal length, the defined reference plane is the anterior corneal surface z Calculated using the aqueous index of refraction (n Z 1.336); for the purpose of determining the focal length, the defined reference plane is also the anterior corneal surface x Calculated using the aqueous index of refraction (n Z 1.336); for the purpose of determining the focal length, the defined reference plane is the posterior corneal surface

Outcomes with Simulated Keratometry

Table 2 shows the results of IOL power calculation using simulated K. In the whole group, all formulas achieved good results with a prediction error of G0.50 D or less in at least 90 (76.3%) of 118 eyes and a median absolute error less than 0.30 D. The highest percentage (100 [84.75%] of 118 eyes) with a prediction error of G0.50 D or less and the lowest median absolute error were obtained by the Barrett Universal II formula. However, Friedman’s test did not reveal any statistically significant difference (P Z .1083) among the tested formulas. Figure 1 shows the distribution of the prediction error for each formula and visually confirms that no major differences could be observed among the different formulas. In the Italian group, slightly better results were obtained with the Barrett Universal II and Haigis formulas. The former yielded the lowest median absolute error and both, the highest percentage (57 [83.82%] of 68) with a prediction error of 0.50 D or less. However, no statistically significant differences were detected between the 5 formulas. In the Japanese group, the best results were obtained with the Barrett Universal II and SRK/T formulas, which yielded the highest percentage (43 [86.00%] of 50 eyes and 41 [82.00%] of 50 eyes, respectively) with a prediction error of G0.50 D or less and the lowest median absolute error. According to Friedman’s test, the difference in absolute prediction error between the 5 formulas was statistically significant (P Z .0012); Dunn’s post test revealed that the only significant difference was between SRK/T and both the Haigis and Hoffer Q formulas (P ! .05). In the whole sample, the optimized constants of the Hoffer Q, Holladay 1, and SRK/T formulas, shown in Table 2, were higher than those reported on the ULIB website. The difference was 0.18 for the Hoffer Q (5.82 versus 5.64), 0.18 for the Holladay 1 (2.02 versus 1.84), and 0.26 for the SRK/T (119.26 versus 119.00). A similar difference was observed for the Barrett Universal II, whose optimized constant was higher by 0.10 with respect to that provided on the website (2.08 versus 1.98). The optimized constants of the Haigis formula were slightly lower (a0 and a1) or

higher (a2) than those reported on the ULIB website. Entering the ULIB constants would have produced an average hyperopic prediction error ranging between 0.17 D and 0.19 D, an increase in the median absolute error ranging between 0.01 D (Hoffer Q and SRK/T) and 0.05 D (Haigis), and would have only minimally affected the percentage of eyes with an absolute prediction error of 0.50 D or less. In the comparison between the Italian and the Japanese samples, there were no statistically significant differences in the absolute prediction error of the Barrett Universal II, Hoffer Q, Holladay 1, and SRK/T formulas, although a trend toward slightly better results in the Italian eyes was observed with most formulas. From a statistical point of view, the only exception was the Haigis formula, which yielded a significantly lower absolute prediction error in the Italian sample (P Z .0321). Outcomes with Total Corneal Power

Table 3 shows the results of IOL power calculation using the 3 TCP values. After constant optimization, the 3 TCP values led to similar outcomes when entered into the third-generation formulas. When the results of the Haigis, Hoffer Q, Holladay 1, and SRK/T formulas were averaged, 89.5 cases, 87.0 cases, and 89.75 cases had a prediction error of G0.50 D or less with TCP1, TCP2, and TCP-IOL, respectively. When used to compare TCP1, TCP2, and TCP-IOL, Friedman’s test did not reveal any statistically significant difference in the absolute prediction error with the Haigis (P Z .7907), Hoffer Q (P Z .1234), and Holladay 1 (P Z .0569) formulas. With the SRK/T formula, by contrast, there was a significant difference (P Z .0005) and it was related to worse results with TCP 2 as compared with TCP 1 and TCP IOL. Comparison Between Total Corneal Power and Simulated Keratometry

Entering TCP values into third-generation formulas did not lead to better refractive results than those obtained Volume - Issue - - 2018

4

IOL POWER CALCULATION WITH A NEW BIOMETER

Table 2. Results of IOL power calculation using simulated K. Group/Formula Sim-K (whole group) Haigis19

Hoffer Q20,21 Holladay 122 SRK/T23 Average of the 4 formulas above Barrett Universal IIB Sim-K (Italian group) Haigis19

Hoffer Q20,21 Holladay 122 SRK/T23 Average of the 4 formulas above Barrett Universal IIB Sim-K (Japanese group) Haigis19

Hoffer Q20,21 Holladay 122 SRK/T23 Average of 4 formulas above Barrett Universal IIB

Optimized Constant

PE (D)

Variance (D)

MedAE (D)

MAE (D)

% Eyes w/PE % ±0.50 D

a0 Z 0.8805 a1 Z 0.1828 a2 Z 0.2345 5.82 2.02 119.26

0.01

0.18

0.28

0.34

76.27

0.05 0.04 0.02

0.19 0.17 0.14

0.26 0.29 0.27

0.35 0.33 0.30

2.08

0.00

0.15

0.22

0.29

77.12 78.81 82.20 78.60 84.75

a0 Z 0.0961 a1 Z 0.2204 a2 Z 0.1936 5.73 1.94 119.18

0.00

0.14

0.25

0.29

83.82

0.01 0.01 0.00

0.14 0.14 0.15

0.26 0.23 0.25

0.30 0.29 0.30

2.02

0.01

0.13

0.19

0.26

83.82 80.88 83.82 82.35 83.82

a0 Z 0.5400 a1 Z 0.234 a2 Z 0.217 5.90 2.14 119.32

0.01

0.23

0.34

0.39

76.00

0.03 0.00 0.00

0.26 0.20 0.14

0.32 0.31 0.26

0.40 0.34 0.30

2.12

0.04

0.14

0.24

0.30

72.00 78.00 82.00 77.00 86.00

IOL Z intraocular lens; K Z keratometry; MAE Z mean absolute error; MedAE Z median absolute error; PE Z prediction error; Sim-K Z simulated keratometry

with simulated K. As regards the absolute prediction error, third-generation formulas did not show any statistically significant difference between the values obtained with simulated K and any TCP value, either in the whole group or in the 2 subgroups. In the whole group, simulated K enabled third-generation formulas to collectively obtain a prediction error of G0.50 D or less in 92.75 (78.60%) of 118 eyes (Table 2). This percentage was slightly higher than those reported above for TCP1, TCP2, and TCP-IOL (ranging between 87.0 eyes [73.73%] and 89.75 eyes [76.06%]).

DISCUSSION Our data show that the biometric data provided by the Scheimpflug–Placido optical biometer can be used to accurately calculate the IOL power in eyes having cataract surgery. In the whole sample, the percentage of eyes with a prediction error of G0.50 D or less ranged between 72.03% and 84.00%, depending on the corneal power and the formula used. These percentages are well above the 55% value established as the benchmark by the National Health Service of the United Kingdom.29 Slightly better outcomes were achieved with simulated K (as opposed to

Figure 1. Distribution of the prediction error with the tested formulas across the whole sample. The central box represents the values from the lower to upper quartile (25th to 75th percentile). The middle line represents the median absolute error (UII Z Universal II).

Volume - Issue - - 2018

5

IOL POWER CALCULATION WITH A NEW BIOMETER

Table 3. Results of IOL power calculation using TCP1, TCP2, and TCP-IOL. TCP/Formula TCP1 (Whole Group)* Haigis19

Hoffer Q20,21 Holladay 122 SRK/T23 Average of 4 formulas above TCP2 (Whole Group)† Haigis19

Hoffer Q20,21 Holladay 122 SRK/T23 Average of 4 formulas above TCP-IOL (Whole Group)z Haigis19

Hoffer Q20,21 Holladay 122 SRK/T23 Average of 4 formulas above

Optimized Constant

Variance (D)

MedAE (D)

MAE (D)

% Eyes with PE % ±0.50 D

0.02

0.22

0.29

0.36

74.58

0.02 0.01 0.01

0.21 0.19 0.18

0.27 0.25 0.30

0.35 0.34 0.34

72.88 75.42 80.51 75.85

a0 Z 3.221 a1 Z 0.1394 a2 Z 0.0105 4.42 0.77 117.47

0.01

0.20

0.26

0.34

75.42

0.05 0.04 0.06

0.23 0.22 0.25

0.27 0.28 0.30

0.36 0.36 0.38

72.88 74.58 72.03 73.73

a0 Z 1.964 a1 Z 0.1211 a2 Z 0.0892 4.98 1.26 118.19

0.00

0.22

0.27

0.35

75.42

0.01 0.01 0.03

0.21 0.21 0.20

0.24 0.24 0.27

0.34 0.33 0.34

74.58 76.27 77.97 76.06

a0 Z 0.3843 a1 Z 0.1301 a2 Z 0.1697 5.36 1.61 118.68

PE (D)

IOL Z intraocular lens; MAE Z mean absolute error; MedAE Z median absolute error; PE Z prediction error; TCP Z total corneal power *Calculated using the corneal index of refraction (n Z 1.376); for the purpose of determining the focal length, the defined reference plane is the anterior corneal surface † Calculated using the aqueous index of refraction (n Z 1.336); for the purpose of determining the focal length, the defined reference plane is also the anterior corneal surface z Calculated using the aqueous index of refraction (n Z 1.336); for the purpose of determining the focal length, the defined reference plane is the posterior corneal surface

TCP), which enabled the different formulas to achieve a median absolute error between 0.22 D and 0.29 D and a percentage of eyes with a prediction error equal to or less than 0.50 D between 76.27% and 84.00%. These values are close to those previously reported with other optical biometers (Table 4), as well as to those reported by Ventura et al.13 with the same device on a smaller sample with the Haigis formula.19 It should be noted that the optimized constants of the Hoffer Q,20,21 Holladay 1,22 and SRK/T23 formulas are higher than those published on the ULIB websiteB for the IOLMaster and the same IOL model. The same comment is also valid for the Barrett Universal II formula and the Asia-Pacific Association of Cataract and Refractive Surgeons website.B If the ULIB constants had been used, an average hyperopic error (approximately 0.18 D) would have occurred with all formulas. Hence, to take full advantage of the technology of this device, we recommend using specifically optimized constants. Regarding TCP, 2 findings deserve our attention. First, the results for IOL power calculation were similar to those obtained by means of simulated K because no statistically significant differences were detected between TCP-based and simulated K-based calculations (similar results have been recently reported by our group when comparing the

TCP-based and Sim-K based values from another Scheimpflug camera).30 Second, no significant differences were detected between the 3 TCP modalities available on the device. Therefore, although simulated K offers slightly better outcomes, we can conclude that all corneal power measurements by the Scheimpflug–Placido optical biometer can be used to calculate the IOL power. A mandatory condition for using any TCP for IOL power calculation is constant optimization to compensate for the systematic difference in the calculated corneal power, which was lowest with TCP2. The differences can be substantial; for example, the A-constant of the SRK/T formula can range from 117.47 (TCP-2) to 119.26 (simulated K). Most formulas yielded the same accuracy in white and Asian eyes, as the prediction error did not show any statistically significant difference with the Barrett Universal II,B Hoffer Q,20,21 Holladay 1,22 and SRK/T23 formulas (the only exception was the Haigis19 formula). However, we could obtain these results only by separately optimizing the constants for the 2 populations. All lens constants were higher in the Japanese group: for example, the optimized A-constant was 119.18 and 119.32 in Italian and Japanese eyes, respectively. This difference, which is close to that also reported on the ULIB website for the IOLMaster,A is likely to depend on the racial differences of biometric parameters, as recently Volume - Issue - - 2018

6

IOL POWER CALCULATION WITH A NEW BIOMETER

Table 4. Accuracy of biometric measurements for IOL power calculation with different optical biometers and formulas (only the results of the best formula for each study are shown). Instrument

IOL Model

Patients (N)

Formula

MedAE (D)

MAE (D)

% Eyes with PE % ±0.50 D

Present study

Galilei G6

118

BarrettB

0.22

0.29

84.75

Ventura 201713

Galilei G6

Acrysof SN60WF Acrysof SN60WF Acrysof SN60WF Acrysof SA60AT Acrysof SN60WF NA AcrySof MA60AC PY-60AD (Hoya) Acrysof SN60WF Acrysof SN60WF Acrysof SN60WF Acrysof SN60WF 3 IOL modelsz Acrysof SN60WF Acrysof SN60WF

37

Haigis19

0.26

NA

86.49

74

Hoffer Q20,21

0.25

0.27

89.04

59

Holladay 122

0.29

0.32

81.45

1079

BarrettB

0.25

0.31

80.6

77 461

Holladay 122 Olsen4

NA NA

0.44 0.43

NA 62.5

163

Hoffer Q20,21

0.32

0.40

68.0†

3241

BarrettB

0.30

0.38

72.3

50

Haigis19

NA

0.46

56.0

1079

Olsen4

0.22

0.28

83.7

50

Haigis19

NA

0.45

58.0

308 82

Holladay 122 Holladay 122

0.26 0.25

0.31 0.31

79.2 NA

249

Hoffer Q20,21

0.34

0.39

71.5

Study*

Savini 201712

Aladdin

Hoffer 20169

IOLMaster 700

Cooke 20162

IOLMaster

Findl 20013 Olsen 20074

IOLMaster IOLMaster

Srivannaboon 20136

IOLMaster

Kane 20167

IOLMaster

Hoffer 20108

IOLMaster

Cooke 20162

Lenstar

Hoffer 20108

Lenstar

Hoffmann 201310 Hoffmann 201310

Lenstar Lenstar

Savini 201716

OA-2000

IOL Z intraocular lens; MAE Z mean absolute error; MedAE Z median absolute error; NA Z not available; PE Z prediction error *First author † The percentage refers only to eyes with an axial length between 22.0 mm and 24.5 mm z Acrysof SN60WF (Alcon Laboratories, Inc.), iMics1 (Bausch & Lomb, Inc.), Tecnis (Abbott Medical Optics, Inc.)

highlighted by Hoffer and Savini.31,32 Hence, race is one more reason for all users of any optical biometer to back-calculate the optimized constants in their patients in order to obtain the best refractive outcomes. The results of the present study were for some reason worse than those previously reported by our own group using keratometry with the Scheimpflug–Placido optical biometer and AL measured by immersion ultrasound (US) biometry,33 which enabled the Hoffer Q,20,21 Holladay 1,22 and SRK/T23 to yield an MAE of 0.21 D, 0.22 D, and 0.25 D, respectively (as compared with an MAE of 0.30 D, 0.29 D, and 0.30 D using the same formulas in the Italian subsample of the current study). A comparison of the median absolute error (not published in our previous paper) also showed better results in 2011 than in 2017, as it increased from 0.16 to 0.26 with the Hoffer Q,20,21 from 0.18 to 0.23 with the Holladay 1,22 and from 0.17 to 0.25 with the SRK/T.23 One possible explanation was that in our previous study, AL was measured by US immersion biometry rather than optical biometry and to rule out this possibility, we analyzed the results in the Italian group by substituting optical (mean value 24.33 G 1.13 mm) with Volume - Issue - - 2018

US biometry (mean value 24.27 G 1.18 mm) in the 40 eyes in which AL had been measured by both techniques. The median absolute error did not improve as a consequence of shifting from optical biometry to US immersion biometry. Because the sample size was similar in the 2 groups (40 versus 43 eyes) and no significant differences were observed for mean simulated K and AL values, we can hypothesize that the better outcomes with the Scheimpflug–Placido optical biometer are related to the implantation of 3-piece IOLs in our previous study instead of the 1-piece IOLs used in this study; 3-piece IOLs, in fact, have been shown to lead to better refractive outcomes than 1-piece IOLs.34 This study has some limitations that warrant further investigation. A direct comparison to other optical biometers was not carried out; however, the literature provides us with a large number of these studies and Table 4 can easily be used for this purpose. Some formulas, such as Olsen and Hoffmann’s,35 were not evaluated, but these are not included in the Scheimpflug– Placido optical biometer internal software and unlike the Barrett Universal II formula, are not available on

IOL POWER CALCULATION WITH A NEW BIOMETER

the Internet. On the other hand, specific ray-tracing software (Okulix, Tomey Corp.) was available for the Italian subsample, but not for the Japanese one, and led to worse results compared with the standard formulas (median absolute error Z 0.34 D, eyes with prediction error equal to or less than 0.50 D Z 59.62%). Finally, we did not separately analyze the results in the different subsamples of eyes based on AL (ie, short, medium, medium-long, and long eyes), because the number of short (n Z 3) and long (n Z 12) eyes was not sufficient. In conclusion, we found that the Scheimpflug–Placido optical biometer offers accurate measurements for IOL power calculation in unoperated eyes. Any corneal power measurement, either based on simulated K or corneal ray tracing, can be used on the condition that the appropriate formula optimized constants are used.

WHAT WAS KNOWN  Biometric measurements provided by earlier optical biometers (IOLMaster and Lenstar) lead to accurate calculations of the IOL power.  Biometric measurements by a newer optical biometer (Galilei G6) combined with a dual Scheimpflug analyzer and a Placido topographer show good agreement with respect to those obtained with the IOLMaster.

WHAT THIS PAPER ADDS  Biometric measurements with the Scheimpflug–Placido optical biometer led to accurate IOL power calculation when entered into the Barrett Universal II, Haigis, Hoffer Q, Holladay 1, and SRK/T formulas.  Using simulated K and TCP by ray tracing led to similar outcomes for IOL power calculation on the condition that the formula constants were optimized.  Racial differences had an effect on lens constants.

REFERENCES 1. Aristodemou P, Knox Cartwright NE, Sparrow JM, Johnston RL. Formula choice: Hoffer Q, Holladay 1, or SRK/T and refractive outcomes in 8108 eyes after cataract surgery with biometry by partial coherence interferometry. J Cataract Refract Surg 2011; 37:63–71 2. Cooke DL, Cooke TL. Comparison of 9 intraocular lens power calculation formulas. J Cataract Refract Surg 2016; 42:1157–1164 3. Findl O, Drexler W, Menapace R, Heinzl H, Hitzenberger CK, Fercher AF. Improved prediction of intraocular lens power using partial coherence interferometry. J Cataract Refract Surg 2001; 27:861–867 4. Olsen T. Improved accuracy of intraocular lens power calculation with the Zeiss IOLMaster. Acta Ophthalmol Scand 2007; 85:84–87. Available at: https://onlinelibrary.wiley.com/doi/epdf/10.1111/j.1600-0420.2006.00774 .x. Accessed April 20, 2018 5. Reitblat O, Assia EI, Kleinmann G, Levy A, Barrett GD, Abulafia A. Accuracy of predicted refraction with multifocal intraocular lenses using two biometry measurement devices and multiple intraocular lens power calculation formulas. Clin Exp Ophthalmol 2015; 43:328–334. Available at: http: //onlinelibrary.wiley.com/doi/10.1111/ceo.12478/pdf. Accessed April 20, 2018 6. Srivannaboon S, Chirapapaisan C, Chirapapaisan N, Lertsuwanroj B, Chongchareon M. Accuracy of Holladay 2 formula using IOLMaster parameters in the absence of lens thickness value. Graefes Arch Clin Exp Ophthalmol 2013; 251:2563–2567 7. Kane JX, Van Heerden A, Atik A, Petsoglou C. Intraocular lens power formula accuracy: comparison of 7 formulas. J Cataract Refract Surg 2016; 42:1490–1500

7

8. Hoffer KJ, Shammas HJ, Savini G. Comparison of 2 instruments for measuring axial length. J Cataract Refract Surg 2010; 36:644–648; erratum, 1066 9. Hoffer KJ, Hoffmann PC, Savini G. Comparison of a new optical biometer using swept-source optical coherence tomography and a biometer using optical low-coherence reflectometry. J Cataract Refract Surg 2016; 42:1165–1172 10. Hoffmann PC, Lindemann CR. Intraocular lens calculation for aspheric intraocular lenses. J Cataract Refract Surg 2013; 39:867–872 11. Olsen T. Use of fellow eye data in the calculation of intraocular lens power for the second eye. Ophthalmology 2011; 118:1710–1715 12. Savini G, Hoffer KJ, Barboni P, Balducci N, Schiano-Lomoriello D, Ducoli P. Accuracy of optical biometry combined with Placido disc corneal topography for intraocular lens power calculation. PLoS One 2017; 12:e0172634. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles /PMC5322924/pdf/pone.0172634.pdf. Accessed April 20, 2018 13. Ventura BV, Ventura MC, Wang L, Koch DD, Weikert MP. Comparison of biometry and intraocular lens power calculation performed by a new optical biometry device and a reference biometer. J Cataract Refract 2017; 43:74–79 14. Kaya F, Kocak I, Aydin A, Baybora H, Karabela Y. Comparison of different formulas for intraocular lens power calculation using a new optical biometer. J Fr Ophtalmol 2015; 38:717–722 15. Kaswin G, Rousseau A, Mgarrech M, Barreau E, Labetoulle M. Biometry and intraocular lens power calculation results with a new optical biometry device: comparison with the gold standard. J Cataract Refract Surg 2014; 40:593–600 16. Savini G, Hoffer KJ, Shammas HJ, Aramberri J, Huang J, Barboni P. Accuracy of a new swept-source optical coherence tomography biometer for IOL power calculation and comparison to IOLMaster. J Refract Surg 2017; 33:690–695 17. Chylack LT Jr, Wolfe JK, Singer DM, Leske MC, Bullimore MA, Bailey IL, Friend J, McCarthy D, Wu S-Y, for the Longitudinal Study of Cataract Study Group. The Lens Opacities Classification System III. Arch Ophthalmol 1993; 111:831–836; erratum 1506 18. Hoffer KJ, Aramberri J, Haigis W, Olsen T, Savini G, Shammas HJ, Bentow S. Protocols for studies of intraocular lens formula accuracy. Am J Ophthalmol 2015; 160:403–405 19. Haigis W, Lege B, Miller N, Schneider B. Comparison of immersion ultrasound biometry and partial coherence interferometry for intraocular lens calculation according to Haigis. Graefes Arch Clin Exp Ophthalmol 2000; 238:765–773 20. Hoffer KJ. The Hoffer Q formula: a comparison of theoretic and regression formulas. J Cataract Refract Surg 1993; 19:700–712; errata, 1994; 20:677; 2007; 33:2–3 21. Zuberbuhler B, Morrell AJ. Errata in printed Hoffer Q formula [letter]. J Cataract Refract Surg 2007; 33:2; reply by KJ Hoffer, 2–3 22. Holladay JT, Prager TC, Chandler TY, Musgrove KH, Lewis JW, Ruiz RS. A three-part system for refining intraocular lens power calculations. J Cataract Refract Surg 1988; 14:17–24 23. Retzlaff JA, Sanders DR, Kraff MC. Development of the SRK/T intraocular lens power calculation formula. J Cataract Refract Surg 1990; 16:333– 340; erratum, 528 24. Masket S, Tennen DG. Astigmatic stabilization of 3.0 mm temporal clear corneal cataract incisions. J Cataract Refract Surg 1996; 22:1451–1455 25. Simpson MJ, Charman WN. The effect of testing distance on intraocular lens power calculation [letter]. J Refract Surg 2014; 30:726. Available at: https://www.healio.com/ophthalmology/journals/jrs/2014-11-30-11/%7B 2806aae9-a03c-4ed8-8e61-f9246494f8fa%7D/the-effect-of-testing-dis tance-on-intraocular-lens-power-calculation. Accessed April 20, 2018 26. Hoffer KJ. Accuracy of ultrasound intraocular lens calculation. Arch Ophthalmol 1981; 99:1819–1923 27. Hoffer KJ. Clinical results using the Holladay 2 intraocular lens power formula. J Cataract Refract Surg 2000; 26:1233–1237 28. Savini G, Carbonelli M, Barboni P, Hoffer KJ. Repeatability of automatic measurements performed by a dual Scheimpflug analyzer in unoperated and post-refractive surgery eyes. J Cataract Refract Surg 2011; 37:302–309 29. Gale RP, Saldana M, Johnston RL, Zuberbuhler B, McKibbin M. Benchmark standards for refractive outcomes after NHS cataract surgery. Eye 2009; 23:149–152. Available at: http://www.nature.com/eye/journal/v23 /n1/pdf/6702954a.pdf. Accessed April 20, 2018 30. Savini G, Hoffer KJ, Schiano Lomoriello D, Ducoli P. Simulated keratometry versus total corneal power by ray-tracing: a comparison in prediction accuracy of intraocular lens power. Cornea 2017; 36:1368–1372 31. Hoffer KJ, Savini G. Effect of gender and race on ocular biometry. Int Ophthalmol Clin 2017; 57(3):137–142 32. Hoffer KJ, Savini G. Clinical results of the Hoffer H-5 formula in 2707 eyes: first 5th-generation formula based on gender and race. Int Ophthalmol Clin 2017; 57(4):213–219

Volume - Issue - - 2018

8

IOL POWER CALCULATION WITH A NEW BIOMETER

33. Savini G, Barboni P, Carbonelli M, Hoffer KJ. Accuracy of a dual Scheimpflug analyzer and a corneal topography system for intraocular lens power calculation in unoperated eyes. J Cataract Refract Surg 2011; 37:72–76 34. Savini G, Barboni P, Ducoli P, Borrelli E, Hoffer KJ. Influence of intraocular lens haptic design on refractive error. J Cataract Refract Surg 2014; 40:1473–1478 35. Olsen T, Hoffmann P. C constant: new concept for ray tracing–assisted intraocular lens power calculation. J Cataract Refract Surg 2014; 40:764–773. Available at: http://www.jcrsjournal.org/article/S0886-3350 (14)00249-1/pdf. Accessed April 20, 2018 OTHER CITED MATERIAL A. User Group for Laser Interference Biometry. ULIB Support for the Zeiss IOLMaster. Available at: http://ocusoft.de/ulib/czm/index.htm. Accessed April 20, 2018

Volume - Issue - - 2018

B. Barrett GD. Barrett Universal II Formula. Singapore, Asia-Pacific Association of Cataract and Refractive Surgeons. Available at: http://www.apacrs .org/barrett_universal2/. Accessed April 20, 2018

Disclosures: Dr. Savini is a consultant to Sifi Medtech Srl. Dr. Hoffer licenses the registered trademark name Hoffer Ò to ensure accurate programming of his formulas to Carl Zeiss Meditec AG €te (IOLMaster), Haag-Streit AG (Lenstar), Oculus Optikgera GmbH (Pentacam AXL), Movu, Inc. (Argos), Nidek Co. Ltd. (ALScan), Tomey Corp. (OA-2000), Topcon Europe Medical B.V./Visia Imaging S.r.l. (Aladdin), Ziemer Ophthalmology GmbH (Galilei G6), and all A-scan biometer manufacturers. None of the other authors has a financial or proprietary interest in any material or method mentioned.