Performance of three biometry devices in patients with different grades of age-related cataract

Authors


Michael Georgopoulos, MD, MBA
Department of Ophthalmology
Medical University of Vienna
Währinger Gürtel 18-20
1090 Vienna
Austria
Tel: + 43-1-40400-4520
Fax: + 43-1-40400-7932
Email: michael.georgopoulos@meduniwien.ac.at

Abstract.

Purpose:  The new Lenstar biometry device was compared in a typical clinical setting to the IOL-Master and Visante-OCT.

Methods:  Fifty-one eyes of 51 patients with age-related cataract were examined with Lenstar LS900 (Haag Streit AG) biometer, IOL-Master V.5 (Carl Zeiss Meditec AG) and Visante-OCT (Carl Zeiss Meditec AG) before cataract surgery. Central corneal thickness (CCT), anterior chamber depth (ACD), keratometry readings of flattest and steepest meridian (K), corneal radius (R) and axial length (AL) values were correlated. Cataracts were graded according to the Lens Opacities Classification System III (LOCS) regarding nuclear colour (NC), nuclear opalescence (NO), cortical (C) and posterior subcapsular (P) cataract.

Results:  Mean values and standard deviations for AL, K and R was 23.66 ± 1.23 mm and 23.67 ± 1.26 mm, 43.24 ± 1.69 dpt and 43.16 ± 1.71 dpt, 7.68 ± 0.29 mm and 7.70 ± 0.28 mm with the IOL-Master and with the LS900, respectively (r = 0.99 and p = 0.76, r = 0.99 and p = 0.029, r = 0.89 and p = 0.14, respectively). Visante-OCT demonstrated highest values of three devices regarding to ACD followed by Lenstar LS900 and IOLMaster. Axial length measurements were unfeasible in 10% of the cases (five patients) and this significantly correlated with the presence of posterior subcapsular cataract of LOCS III grade 4.0 or higher.

Conclusions:  IOL-Master, Lenstar LS900 and AC–OCT proved to be excellent non-contact measurement methods in eyes with age-related cataract. Nevertheless, ultrasound biometry is still required for cases with dense posterior subcapsular cataract.

Introduction

Ultrasound biometry has been the gold standard for many years and provides measurements of the corneal thickness (CT), anterior chamber depth (ACD) and axial length (AL). It is well known that due to indentation of the cornea during the measurement, the results of this method are variable and also depend on the exact axial placement of the probe relative to the centre of the cornea (Solomon 1999; Marsich & Bullimore 2000; Miglior et al. 2004).

Non-contact devices are easy to use, quick and offer significantly higher resolution of axial length measurements (Rainer et al. 2004). Thus, currently the non-contact methods are preferred for biometry of the eye (Rainer et al. 2002). The partial coherence interferometry (IOL-Master, Carl Zeiss Meditec AG, Jena, Germany) is supposed to be the most frequently used biometry device in current clinical practice. However, no AL measurements could be performed in eyes with advanced cataracts (Freeman & Pesudovs 2005; Suto et al. 2007; Hill et al. 2008).

Another technique, which allows high resolution visualization of the anterior segment using optical coherence tomography, is the Visante-OCT (Carl Zeiss Meditec AG).

Recently, a novel ocular biometry device was developed by Haag-Streit (Lenstar LS900, Haag-Streit Koeniz, Switzerland). The Lenstar LS900 (LS900) provides several different measurements (Fig. 1).

Figure 1.

 Parameters evaluated by Lenstar LS900 (Haag Streit AG) device: axial length (AL), anterior chamber depth (ACD), anatomical anterior chamber depth (AD), lens thickness (LT), central corneal thickness (CCT), retinal thickness (RT).

This study was conducted to compare three different biometry devices and to evaluate the measurement performance of IOL-Master, Lenstar LS900 and Visante-OCT in patients with different stages of cataract classified using the Lens Opacities Classification System (LOCS).

Materials and Methods

Patient inclusion

Fifty-one eyes of 51 consecutive patients were included in this observational, cross-sectional clinical study in July 2009. All patients were scheduled for cataract surgery and exclusion criteria were moderate and severe corneal or macular diseases.

Following a standardized ophthalmologic examination procedure comprising Snellen visual acuity test, dilated pupil ophthalmoscopy at the slitlamp, non-contact biometry with IOL-Master V.5, LS900 biometer and Visante-OCT imaging (model 1000, software version 1.0, Carl-Zeiss Meditec) was performed. Due to advanced cataract severity in five patients the IOL-Master did not provide AL measurements and therefore standard A-scan applanation ultrasound biometry (Ocuscan, Alcon) was also performed in these eyes.

All examinations were conducted at the Department of Ophthalmology, Medical University of Vienna, Austria. All measurements were performed after the purpose of the study was explained and informed consent given.

Lens Opacities Classification System

Type and severity of cataract were evaluated using LOCS III grading system, evaluating the opacity of the lens with respect to nuclear colour (NC, score 0 to 6), nuclear opalescence (NO, score 0 to 6), cortical cataract (C, score 0 to 5) and posterior subcapsular (P, score 0 to 5) cataract (Chylack et al. 1993). Each parameter was assessed subjectively at the slitlamp by one examiner and documented in the case report forms.

Biometry

Non-contact biometric measurements using IOL-Master, LS900 and Visante-OCT were performed in all patients following a standardized order. All measurements were performed at the same day and by the same examiner. If it was not possible to perform AL measurement with the IOL-Master, an A-scan applanation ultrasound was also performed.

IOL-Master

The IOL-Master is a non-contact optical device that measures the distance from the corneal vertex to the retinal pigment epithelium by partial coherence interferometry. Four axial length, three keratometry and five ACD measurements were performed with the IOL-Master and the mean measurement results were used.

Lenstar LS900

The LS900 is a new ocular biometry device provides several different measurements including corneal thickness, ACD, lens thickness, AL, keratometry, white-to-white distance, pupillometry, eccentricity of the visual optical line and central retinal thickness. The length measurements are performed using optical coherence technology with a super luminescent diode as light source. Measurement of axial length along the visual optical axis is performed while the device controls the stable fixation. Four complete measurements were performed in each eye and analysed with the LS900 device.

Visante-OCT

The Visante-OCT achieves an axial resolution of approximately 18 μm and provides highly detailed, in-depth images of the anterior chamber and measures parameters such as corneal thickness, ACD and anterior chamber angle. The resolution ranges from 128 to 512 A-scans per line. For the purpose of our study three OCT scans were acquired using the anterior segment single 0° to 180° protocol (8 mm deep by 16 mm wide, with 256 A-scans per line) and one high-resolution corneal scan (10 mm deep by 3 mm wide, with 512 A-scans per line). ACD was measured from corneal endothelium to crystalline lens using the chamber function as provided by the manufacturer.

Central corneal thickness (CCT) was measured along a line perpendicular to the anterior and posterior surface of the cornea in the zone of the optically produced corneal reflex. A single high-resolution corneal scan was used as CCT in this study. Central ACD (between the anterior surface of the lens to the anterior surface of the cornea) and anatomical ACD (between anterior surface of the lens to the posterior surface of the cornea) were evaluated using the anterior segment images. The mean of three readings was calculated for comparison of ACD measurements with the other devices.

All data were entered into a Microsoft Excel spreadsheet. Bland-Altman plots were used to find a potential dependency between differences and means of three measurements. p values of less than 0.05 were considered to be statistically significant. Statistical analysis and mean values, standard deviation, minimum and maximum were calculated using Microsoft Excel Software.

Results

Demographic data

The mean age of the 51 patients was 68 ± 11 years. Fifty-one eyes and 22 (43%) patients were female. Twenty-six right and 25 left eyes were included. Mean visual acuity was 0.4 ± 0.2 SNELLEN lines (range, HM-0.8).

The mean LOCS grading of the patients was for NO 3.1, for NC 3.0, for C 1.9 and for P 1.1. In seven patients the posterior subcapsular cataract (P) severity was graded ≥4 using LOCS III. In five of these patients it was not possible to perform AL-measurements with the IOL-Master. The LS900 failed to take partial coherence interferometry measurements in these patients and in one additional patient.

Biometry

Mean AL was 23.66 ± 1.23 mm and 23.67 ± 1.26 mm, mean flattest and steepest corneal meridian (R) was 7.68 ± 0.29 mm and 7.70 ± 0.28 mm and mean keratometry readings for flattest and steepest meridian (K) was 43.24 ± 1.69 dpt and 43.16 ± 1.71 dpt with the IOL-Master and with the LS900, respectively. The differences between the results of the two instruments regarding AL, R and K were not statistically significant (r = 0.99 and p = 0.76, r = 0.89 and p = 0.14, r = 0.99 and p = 0.029, respectively). Mean and maximum paired differences, median absolute differences and upper and lower limits of agreement between the two methods are shown in Table 1. Bland-Altman plots of the paired AL, R and K differences against the mean values are shown in Fig. 2.

Table 1.   Axial length (AL), keratometry readings for flattest and steepest meridian (K), corneal radius (R), anterior chamber depth (ACD), anatomical anterior chamber depth (AD) lens thickness, central corneal thickness (CCT) measured with all biometry devices.
IOL-MasterMeanSDMinMax
  1. SD = standard deviation, min = minimum, max = maximum.

AL (mm)23.661.2320.7428.13
R (mm)7.680.296.478.22
K (dpt)43.241.6940.3951.48
ACD (mm)3.150.422.234.12
LS900
 AL (mm)23.671.2620.7328.19
 R (mm)7.70.286.478.23
 K (dpt)43.161.7140.3551.45
 CCT (μm)56032425649
 AD (mm) (without CCT) 2.650.381.83.7
 ACD (mm)3.210.392.44.25
 Lens thickness (mm)4.570.423.765.62
Visante-OCT
 AD (mm) (without CCT)2.760.371.933.79
 ACD (mm)3.320.372.524.34
 CCT (μm)55031433640
 CCT (μm) (high resolution) 55432429640
Figure 2.

 Bland-Altman plots comparing mean and difference axial length (AL), keratometry readings (K) and corneal radius (R) values between IOL-Master and Lenstar LS900 devices.

ACD and CCT

Visante-OCT measured the highest ACD values of the three devices, followed by LS900 and IOL-Master. There was no significant difference between IOL-Master and LS900 (r = 0.82, p = 0.09) but between Visante-OCT and the other two methods, with Visante-OCT always showing higher values (r = 0.80, p < 0.001 compared with IOL-Master and r = 0.98, p < 0.001 compared with LS900). The lowest ACD value was measured by the IOL-Master. Mean ACD for IOL-Master, LS900 and Visante-OCT were 3.15 ± 0.42 mm, 3.21 ± 0.38 mm and 3.32 ± 0.37 mm, respectively.

CCT was 560 ± 32 μm and 554 ± 32 μm with LS900 and with Visante-OCT high-resolution scan (r = 0.97, p < 0.001), respectively.

The Visante-OCT showed higher anatomical ACD (without CCT) values than the LS900. Mean ± SD of anatomical ACD for LS900 and Visante-OCT were 2.65 ± 0.38 mm and 2.76, respectively (r = 0.98, p < 0.001). Bland-Altman plots of the paired ACD, anatomical ACD and CCT differences against the mean values are shown in Fig. 3.

Figure 3.

 Bland-Altman plots comparing mean and difference anterior chamber depth (ACD), central corneal thickness (CCT) and anatomical ACD values between the three devices.

Discussion

In this study we compared the performance of three different biometry devices in patients with different grades of age-related cataract.

Both IOL-Master and LS900 failed to take AL measurements in about 10% of cases (five and six patients, respectively) due to dense media opacities. This percentage compares to the findings of Buckhurst (Buckhurst et al. 2009), however, in their study there was no association with a specific form of cataract. The analysis of cataract severity using the LOCS III grading method showed no influence of the severity of nuclear or cortical cataract (NO, NC or C), but there was a highly significant influence of posterior subcapsular cataract density (P). Problems with axial length measurements significantly correlated with the presence of posterior subcapsular cataract of LOCS III grade 4.0 or higher. In the literature, the severity of subcapsular cataract making axial length measurements impossible with the IOL-Master ranges from 3.5 to 5.0 (Chylack et al. 1993; Suto et al. 2007), however, there are no reports for LS900.

A comparison of partial coherence interferometry with other biometry techniques for precise ocular biometric measurements has been published by several authors (Buckhurst et al. 2009; Holzer et al. 2009; Rohrer et al. 2009). In a previous study with healthy eyes AL and keratometric measurements (R, K) did not differ significantly between IOL-Master and LS900 (Holzer et al. 2009). Buckhurst (Buckhurst et al. 2009) also compared IOL-Master and LS900 in patients with age-related cataract including a subgroup with additional A-scan applanation ultrasonography (OcuScan). They found statistically significant differences between IOL-Master and LS900 regarding AL and ACD values, but rated them as clinically not significant.

Concerning ACD measurements, Visante-OCT showed the highest values, followed by LS900 and IOL-Master. Lavanya (Lavanya et al. 2007) also demonstrated that Visante-OCT measurements of ACD were generally higher than those made with the IOL-Master. This difference might be explained by the fact that Visante-OCT does not offer an automatic mode for ACD measurements and therefore is more investigator dependent than the two automated devices. This may also be the reason for the difference between Visante-OCT and LS900. Another study of Cruysberg et al. (2009) comparing ACD measurements between LS900 and Visante-OCT and also found higher values with Visante-OCT. Our results confirm these results of the previous investigators.

Visante-OCT CCT measurements showed lower values than LS900. It should be noted that the CCT was measured only with one HR scan with Visante-OCT in contrast to four measurements with LS900. This could theoretically influence the results; however, the difference between the two devices was not statistically significant.

In summary, IOL-Master, LS900 and Visante-OCT proved to be excellent non-contact measurement methods in eyes with age-related cataract. Nevertheless, ultrasound biometry is still required for cases with dense posterior subcapsular cataract. Providing measurements for this subgroup of patients will be one of the remaining challenges for the manufacturers of ocular biometry devices.

Ancillary