Anterior chamber depth measurements in eyes with an accommodating intraocular lens

ARTICLE

Anterior chamber depth measurements in eyes with an accommodating intraocular lens Agreement between partial coherence interferometry and optical coherence tomography Georgia Cleary, MB BS, MRCOphth, David J. Spalton, FRCOphth, FRCP, FRCS, John Marshall, PhD, FRCPath, FMedSci

PURPOSE: To determine agreement between partial coherence interferometry (PCI) and anterior segment optical coherence tomography (AS-OCT) measurements of anterior chamber depth (ACD) and axial intraocular lens (IOL) movement in eyes with an accommodating IOL. SETTING: Department of Ophthalmology, St. Thomas’ Hospital, London, United Kingdom. METHODS: In this prospective pilot study of cataract patients, the central ACD was measured by PCI (ACMaster) and AS-OCT (Visante) 6 months after unilateral OPAL-A accommodating IOL implantation. Measurements were performed with a distance target and 1.00 diopter (D) and 2.00 D accommodative targets and after administration of topical pilocarpine 4%. Agreement between PCI and AS-OCT ACD measurements and IOL movement was calculated. RESULTS: Measurements were obtained in 18 patients. There was a consistent and statistically significant bias toward shallower ACD measurements with AS-OCT than with PCI, with the bias most pronounced after pilocarpine (mean 4.117 mm G 0.291 [SD] versus 4.054 G 0.287 mm; bias 0.063 mm; P<.0001). Limited IOL movement to 1.00 D and 2.00 D accommodative stimuli was detected with both instruments. After pilocarpine, forward IOL movement measurements were statistically significantly greater by AS-OCT than by PCI (mean 0.306 G 0.161 mm versus 0.270 G 0.155 mm) (P Z .017). CONCLUSIONS: The AS-OCT device showed a bias toward underestimation of ACD compared with the PCI device. The bias increased as ACD shallowed with pilocarpine, resulting in overestimation of forward IOL movement by AS-OCT. This may lead to overestimation of the accommodative performance of IOLs. The ACD measurements obtained by the 2 devices are not interchangeable. Financial Disclosure: No author has a financial or proprietary interest in any material or method mentioned. J Cataract Refract Surg 2010; 36:790–798 Q 2010 ASCRS and ESCRS

Postoperative spectacle independence is a key aim of contemporary cataract surgery. The ideal outcome for cataract patients would be a continuous range of high-quality vision across all distances from far to near, thus reestablishing the functional status of the prepresbyopic phakic eye. Accommodating intraocular lenses (IOLs) are one of several proposed methods of restoring near vision to pseudophakic eyes. Several accommodating IOL models are commercially available1–6; most use the focus-shift principle. The theoretical mechanism of accommodating IOLs based on the 790

Q 2010 ASCRS and ESCRS Published by Elsevier Inc.

action of focus shift is forward axial movement of the IOL optic produced by ciliary muscle contraction from focusing on a near target, resulting in increased effective IOL power. Supraphysiologic ciliary muscle stimulation can be achieved pharmacologically with topical pilocarpine.2,7 Other commercially available technologies for restoring accommodation in pseudophakic eyes include dual-optic IOLs and capsular bag refilling techniques.8–10 These methods also rely, although to a lesser extent, on movement of the anterior IOL surface. 0886-3350/10/$dsee front matter doi:10.1016/j.jcrs.2009.11.028

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

There are many clinical methods to determine the efficacy of accommodating IOLs. Psychophysical tests of near visual performance are frequently used; these include distance-corrected near visual acuity, subjective amplitude of accommodation, and defocusing with minus lenses. However, these methods are subjective, are influenced by pseudoaccommodative factors, and do not measure true accommodation. To assess a focus-shift IOL, one must measure both objective refractive change and forward movement of the anterior IOL surface, which is measurable as shallowing of the anterior chamber. Accurate measurement of IOL movement is therefore of great practical importance. Open-field aberrometers or autorefractors have been validated as methods of measuring objective changes in refraction in pseudophakic eyes.2,11–13 Various biometric methods have been applied to the measurement of anterior chamber depth (ACD). The most accurate of these is dual-beam partial coherence interferometry (PCI).14–16 Another is anterior segment optical coherence tomography (AS-OCT).18,19 Additional techniques of ACD measurement include traditional A-scan biometry, high-frequency ultrasound biomicroscopy (UBM), scanning-slit topography, and Scheimpflug videokeratography. In this study, we evaluated patients who had implantation of a prototype accommodating IOL that is based on the focus-shift principle. The aim was to determine agreement in ACD measurements and axial IOL movement between a PCI device and an AS-OCT device in eyes with a focus-shift IOL. PATIENTS AND METHODS Patients with cataract were prospectively recruited to a pilot study and had unilateral implantation of an OPAL-A prototype accommodating IOL (HumanOptics AG). The St. Thomas’ Hospital Research Ethics Committee approved the study, and all patients provided written informed consent after receiving an explanation of the nature and possible consequences of the study. The study adhered to the tenets of the Declaration of Helsinki.

Submitted: September 10, 2009. Final revision submitted: November 28, 2009. Accepted: November 30, 2009. From the Departments of Ophthalmology, St. Thomas’ Hospital (Cleary, Spalton) and The Rayne Institute (Marshall), Kings College, London, United Kingdom; and the Centre for Ophthalmology and Visual Science (Cleary), Perth, Western Australia. Presented at the XXVI Congress of the European Society of Cataract & Refractive Surgeons, Berlin, Germany, September 2008. Corresponding author: David J. Spalton, Department of Ophthalmology, St. Thomas’ Hospital, Westminster Bridge Road, London SE1 7EH, United Kingdom. E-mail: [email protected].

791

Measurements Six months postoperatively, the same ophthalmic research fellow performed measurements using a PCI device and an AS-OCT device. The fellow was trained in the use of both devices. The ACD was measured with the 2 instruments in a randomized order, with the patient initially viewing a distance target followed by 1.00 diopter (D) and 2.00 D accommodative targets. Topical pilocarpine 4% was then administered, and the ACD was remeasured 60 minutes later.

Partial Coherence Interferometry

The PCI measurements were taken with an ACMaster device (Carl Zeiss Meditec). This device is a modification of the IOLMaster instrument, which uses PCI for axial length (AL) measurement but a less precise photographic slit-beam technique to measure ACD.17 The ACMaster measures corneal thickness, ACD, and lens thickness using PCI with a longer wavelength than the IOLMaster (850 nm versus 780 nm); measurements are captured parallel to the optical axis of the eye. The calibration of the instrument was checked by following the manufacturer’s instructions using the test eye provided, and subjective refractive data were entered. Patients were instructed to look at the center of the fixation target (an X visible on the liquid crystal display within the instrument) and to follow this target if it moved. The first set of measurements was captured while the patient viewed a distance target. During ACD measurements, 4 Purkinje images were visible as bright reflections on the monitor of the PCI device. The operator superimposed the images by moving the fixation target in the x and y planes. When the images were aligned, a measurement was captured and an axial tracing of the optical scan displayed. Up to 63 axial scans, the maximum permitted by the instrument, were captured parallel to the optical axis. This process was repeated under 1.00 D and 2.00 D of accommodative stimulation by inducing 1.00 D and 2.00 D of defocus, respectively, with the instrument’s built-in optometer; the patient was instructed to continue focusing on the center of the fixation target, keeping it as clear as possible. The procedure was repeated 60 minutes after topical pilocarpine was administered. All scans were later evaluated to ensure their quality. An acceptable tracing was one with 4 clear signal peaks, 1 for each Purkinje image, as follows: anterior corneal surface, posterior corneal surface, anterior lens surface, and posterior lens surface (Figure 1). The PCI device calculates ACD as the distance from the anterior corneal surface to the anterior lens surface. The built-in software accurately labels most anterior and posterior corneal surface signal peaks but is less reliable for anterior and posterior lens surface signal peaks in pseudophakic eyes; thus, these peaks required manual labeling in most cases. Unacceptable traces were deleted, including traces in which the corneal thickness or lens thickness deviated widely from the mean value, suggesting that a different optical path through the anterior segment had been sampled. The instrument calculated a mean value (GSD) for corneal thickness, ACD, and IOL thickness under 4 conditions: distance, 1.00 D accommodation, 2.00 D accommodation, and after pilocarpine administration. Central corneal thickness (CCT) and IOL thickness measurements were used as control measurements to ensure that the same axis was sampled under all conditions. Precision was calculated as the mean of the SDs of distance ACD, corneal thickness, and IOL thickness measurements.

J CATARACT REFRACT SURG - VOL 36, MAY 2010

792

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

Figure 2. An AS-OCT image showing correct ocular alignment. A vertical line running perpendicular to the corneal apex is visible. The anterior chamber tool has been applied to the image for ACD measurement.

Figure 1. A display on the PCI instrument shows an axial tracing obtained in a pseudophakic eye. Four peaks are labeled and represent, from left to right, the anterior corneal surface, posterior corneal surface, anterior lens surface, and posterior lens surface.

Optical Coherence Tomograph The AS-OCT measurements were taken with a Visante OCT instrument (Carl Zeiss Meditec). The OCT images are formed by software algorithms that compensate for distortion due to corneal optical transmission. Distances within the anterior segment are determined semisubjectively from 2-dimensional (2-D) images with various software tools. Calibration was checked using the test eye provided, and patient data (including subjective refraction) were entered. Patients were asked to look at the center of the internal fixation target in the device (a yellow starburst) and were asked to follow the starburst if it moved during measurements. The operator optimized the alignment and position of the anterior segment image on the instrument’s monitor by making adjustments in x, y, and z planes and for angle k by adjusting the angular position of the fixation target. Once the eye was aligned, the eyelids were carefully held open, taking care not to distort the globe. This was necessary because measurements were captured in the quad scan mode, which captures 4 radial images separated by 45 degrees. Each image should include a clear view of the iridocorneal angles, which can be obscured by the eyelid margins, especially superiorly. According to the manufacturer, optimum scan quality is achieved when the scan is captured normal to the corneal vertex. When this occurs, a white line running perpendicular to the corneal vertex is visible on the instrument’s monitor (Figure 2). As with the PCI device, scans were first obtained to determine ACD with the patient viewing a distance target. The procedure was repeated for 1.00 D and 2.00 D accommodative stimuli by introducing defocus with the instrument’s built-in optometer; the patient was asked to keep the center of the fixation target as clear as possible. Last, measurements were taken after pilocarpine administration. Each quad scan generates four 2-D anterior segment images. The built-in image-analysis software includes an anterior chamber tool. This tool was manually applied to each image to determine the ACD from the anterior corneal surface to the anterior lens surface (Figure 2). Four ACD measurements were thus obtained from each quad scan, allowing calculation of mean ACD (GSD) for distance,

1.00 D accommodation, 2.00 D accommodation, and after pilocarpine. The CCT was also measured on each image by the ACD tool. Central IOL thickness was measured by manually applying the device’s caliper tool to the image along the same axis as the ACD, although this was not possible in all eyes because the posterior surface of the IOL was not visible in all OCT images. Central corneal thickness and IOL thickness measurements were recorded to ensure the measurement axis remained constant. Precision was calculated as the mean of the SDs of the 4 distance ACD measurements for each eye and was also calculated for CCT and IOL thickness.

Intraocular Lens Movement For each instrument, IOL movement was calculated as the difference between the distance ACD measurement and the accommodative ACD measurement as follows: IOL movementðaccommodatingÞ Z ACDðdistanceÞ  ACDðaccommodatingÞ Forward movement of the IOL resulted in a positive value and backward movement, in a negative value.

Statistical Analysis Agreement between the 2 instruments was evaluated using the Bland-Altman technique.20 Bias, calculated as the mean of the differences between the 2 devices (PCI minus AS-OCT), and 95% limits of agreement are reported. The statistical significance of differences between measurements was assessed using paired t tests after confirming that the measurements and their differences had a normal distribution. A P value less than 0.05 was considered statistically significant. Linear regression was used to determine the relationship between the dependent variable (IOL movement stimulated by pilocarpine) and 3 independent variables (age, IOL power, AL) for both instruments. GraphPad Prism software (version 5.0b GraphPad Software Inc.) was used.

RESULTS Measurements were taken in 20 of 22 patients; 1 patient died, and another did not attend the follow-up. Measurements with the PCI device were not possible in 2 eyes because the IOL optic was tilted. Thus,

J CATARACT REFRACT SURG - VOL 36, MAY 2010

793

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

Table 1. Anterior chamber depth measurements obtained by PCI and AS-OCT under various accommodative conditions. Mean ACD (mm) G SD Stimulus Distance target 1.0 D accommodation 2.0 D accommodation Pilocarpine

95% LoA*

PCI

AS-OCT

P Value

Bias

95% CI*

Upper Limit

Lower Limit

4.387 G 0.300 4.371 G 0.310 4.364 G 0.311 4.117 G 0.291

4.360 G 0.296 4.345 G 0.311 4.331 G 0.323 4.054 G 0.287

.002 .014 .016 !.0001

0.027 0.026 0.033 0.063

0.012 to 0.041 0.006 to 0.045 0.007 to 0.059 0.038 to 0.087

0.084 0.103 0.135 0.158

0.031 0.052 0.069 0.032

ACD Z anterior chamber depth; AS-OCT Z anterior segment optical coherence tomography; Bias Z mean difference between instruments (PCI  AS-OCT); CI Z confidence interval; LoA Z limits of agreement; PCI Z partial coherence interferometry *Between instruments

complete measurements were obtained in 18 eyes of 18 patients. The mean age of the patients was 62.4 years G 12.5 (SD); 56% were men. The mean IOL power was 22.75 G 2.11 D and the mean AL, 23.17 G 0.74 mm. The mean corrected distance visual acuity was 0.06 G 0.09 logMAR. Table 1 shows the ACD measurements obtained by both devices. The AS-OCT device showed a consistent bias toward smaller ACD measurements than the PCI device (Figures 3 and 4). The difference between the 2 instruments was statistically significant under all test conditions. The largest difference and the highest level of statistical significance were observed when the ACD was measured after pilocarpine stimulation (P!.0001). The 1.00 D and 2.00 D accommodative targets stimulated a small amount of forward IOL movement with both devices (Table 2 and Figures 5 and 6). There was no statistically significant difference in the amount of forward IOL movement between instruments during

near visual stimulation (1.00 D stimuli, P Z .896; 2.0 D stimuli, P Z .477). A larger amount of forward IOL movement was observed after administration of topical pilocarpine. Pilocarpine-stimulated forward IOL movement measured by the AS-OCT device was statistically significantly greater than that measured by the PCI device (P Z .017). Pilocarpine-stimulated IOL movement was not influenced by age or IOL power with either instrument (Table 3). Thirty-two percent of the variation in pilocarpine-stimulated IOL movement was accounted for by AL; a longer

Figure 3. Anterior chamber depth measured under various accommodative conditions. The error bars indicate the SD (* Z P!.05; AS-OCT Z anterior segment optical coherence tomography; PCI Z partial coherence interferometry).

Figure 4. Bland-Altman plots showing agreement between ACD measurements captured viewing a distance target (accommodation relaxed) and after pharmacologic stimulation with pilocarpine (AS-OCT Z anterior segment optical coherence tomography; LoA Z limits of agreement; PCI Z partial coherence interferometry).

J CATARACT REFRACT SURG - VOL 36, MAY 2010

794

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

Table 2. Forward IOL movement measurements obtained by PCI and AS-OCT under various accommodative conditions. Mean Movement (mm) G SD Stimulus 1.0 D accommodation 2.0 D accommodation Pilocarpine

95% LoA*

PCI

AS-OCT

P Value

Bias

95% CI*

Upper Limit

Lower Limit

0.016 G 0.025 0.022 G 0.027 0.270 G 0.155

0.015 G 0.030 0.029 G 0.049 0.306 G 0.161

.896 .477 .017

0.001 0.006 0.036

0.013 to 0.015 0.024 to 0.012 0.065 to -0.007

0.057 0.065 0.077

0.055 0.078 0.149

ACD Z anterior chamber depth; AS-OCT Z anterior segment optical coherence tomography; Bias Z mean difference between instruments (PCI  AS-OCT); CI Z confidence interval; LoA Z limits of agreement; PCI Z partial coherence interferometry *Between instruments

AL was associated with a greater degree of IOL movement. The precision of ACD measurements under distance viewing conditions was 10.6 mm by the PCI device and 13.8 mm by the AC-OCT device. The precision of CCT and IOL thickness measurements by the PCI device under distance viewing conditions was 2.2 mm and 2.5 mm, respectively. The precision of CCT and IOL thickness measurements by the AS-OCT device under distance viewing conditions was 5.4 mm and 15.9 mm, respectively, although IOL thickness data was measurable in 11 eyes only. Complete control data (ie, measurements available under all accommodative conditions) were available for all CCT and IOL thickness measurements by the PCI device (Table 4) and for CCT measurements by the AS-OCT device (Table 5). Complete control data for central IOL thickness measurements with the AS-OCT device were available in 6 eyes because the central posterior IOL surface was not visible on the remaining OCT images under all accommodative conditions.

DISCUSSION

Figure 5. Forward IOL movement under various accommodative conditions. The error bars indicate the SD (* Z P!.05; AS-OCT Z anterior segment optical coherence tomography; IOL Z intraocular lens; PCI Z partial coherence interferometry).

Figure 6. Bland-Altman plots showing agreement between forward IOL movement measurements after pharmacologic stimulation with pilocarpine (AS-OCT Z anterior segment optical coherence tomography; LoA Z limits of agreement; PCI Z partial coherence interferometry).

Accurate ACD measurement is essential for the assessment of axial movement of accommodating IOLs. Many commercially available biometric devices are capable of measuring ACD. In the context of accommodating IOLs, the ideal device would be precise, objective, noncontact, and easy to use and have the ability to alter target distance to stimulate accommodation. The fixation target should be viewed by the eye being measured rather than by the fellow eye to avoid misalignment caused by small amounts of convergence, which may artifactually influence ACD. Although most studies of ACD measurement pertain to phakic eyes, validation of measurement techniques in phakic eyes cannot necessarily be extrapolated to pseudophakic eyes, which have different optical properties including deeper anterior chambers and thinner lenses with a higher refractive index. Few studies have assessed ACD measurement techniques in pseudophakic eyes. Koranyi et al.21 measured ACD 6 weeks postoperatively in 23 eyes with a poly(methyl methacrylate) IOL. The ACD value was significantly less when measured with traditional A-scan biometry (mean 3.73 G 0.26 mm) than with

J CATARACT REFRACT SURG - VOL 36, MAY 2010

795

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

Scheimpflug imaging (mean 4.65 G 0.33 mm). The underestimation of postoperative ACD by A-scan was attributed to the corneal indentation associated with the contact method. Furthermore, the A-scan device was not capable of automatically labeling the IOL peaks. The ACD was remeasured 36 weeks postoperatively in 16 eyes with 4 instruments. The shallowest ACD measurements were again found with A-scan, followed by Scheimpflug imaging, slitlamp-based optical pachymetry, and scanning-slit topography, which yielded the deepest mean ACD and showed the greatest variability. A subsequent study by Nemeth et al.22 compared ACD results obtained with Scheimpflug imaging and A-scan ultrasound in 42 pseudophakic eyes with a normal AL and a hydrophobic acrylic IOL. The authors also found shallower ACD measurements with Scheimpflug imaging than with ultrasound (mean 3.41 G 0.28 mm versus 3.97 G 0.45 mm; P!.001). More recently, 2 UBM devices (HiScan and UBM 840) were used to evaluate anterior segment biometric changes with accommodative effort in a heterogeneous group of eyes with 1 of 4 types of monofocal nonaccommodating IOLs.23 Both devices use an immersion technique for measurement, during which patients were supine and viewed both distance and near targets with the fellow eye. Both instruments detected a small but statistically significant reduction in ACD

with accommodative effort. However, the UBM 840 produced higher estimates of ACD under relaxed (mean 4.26 G 0.43 mm versus 4.09 G 0.41 mm) and active accommodative (mean 4.23 G 0.41 mm versus 4.05 G 0.41 mm) conditions and showed greater variability in its estimation of change in ACD (mean 0.03 G 0.30 mm versus 0.04 G 0.09 mm). Kriechbaum et al.17 compared ACD measurements by the IOLMaster device, which uses high-precision PCI to measure AL but a photographic slit technique to measure ACD, with measurements taken with a laboratory prototype of the ACMaster, which uses PCI to measure ACD. In their study of 34 pseudophakic eyes of 18 patients, the median ACD was 4.50 mm with the PCI measurement and 4.65 mm with photographic slit measurement. The difference between techniques was not statistically significant; however, their correlation was poor (r Z 0.21, PO.2). Several outlying measurements were noted with the photographic slit ACD values, and they were attributed to errors in the device’s imaging technique. A further reason for the discrepancy between the instruments may be that photographic slit measurements were taken after pupil dilation with tropicamide and phenylephrine while PCI measurements were performed with the eye undilated. Cycloplegia may have resulted in ACD deepening in the photographic slit images but would not account for the outlying measurements. The ACMaster was subsequently released commercially and has been shown to be as accurate as the laboratory prototype, with an axial resolution of 6 mm for ACD.15 The ACMaster device we used has many advantages for ACD measurement. These include measurement of ACD in the fixating eye, high precision, good repeatability, and capture of ACD measurements parallel to the optical axis of the eye. Corneal thickness and lens thickness can be assessed in each scan to ensure that the same optical path is sampled during each measurement; however, the device is no longer commercially available and is relatively difficult to use. A substantial period of operator training

Table 4. Central corneal and IOL thickness control measurements by PCI.

Table 5. Central corneal and IOL thickness control measurements by AS-OCT.

Table 3. Relationship between pilocarpine-stimulated IOL movement and age, IOL power, and AL determined by linear regression. PCI Device Parameter Age IOL power Axial length

AS-OCT Device

2

r

P Value

r2

P Value

0.024 0.041 0.315

.538 .418 .015

0.021 0.039 0.318

.565 .430 .015

AS-OCT Z anterior segment optical coherence tomography; IOL Z intraocular lens; PCI Z partial coherence interferometry

Mean (mm) G SD Stimulus Distance target 1.0 D accommodation 2.0 D accommodation Pilocarpine IOL Z intraocular lens

Corneal Thickness (n Z 18)

IOL Thickness (n Z 18)

528.3 G 41.9 527.2 G 42.0 527.4 G 42.1 530.7 G 42.8

836.6 G 66.8 836.8 G 67.1 836.1 G 68.6 837.3 G 65.8

Mean (mm) G SD Stimulus Distance target 1.0 D accommodation 2.0 D accommodation Pilocarpine IOL Z intraocular lens

J CATARACT REFRACT SURG - VOL 36, MAY 2010

Corneal Thickness (n Z 18)

IOL Thickness (n Z 6)

531.9 G 43.4 533.8 G 43.8 531.7 G 43.9 534.9 G 42.8

919.4 G 66.9 921.4 G 68.2 918.6 G 64.5 914.8 G 64.3

796

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

and practice is necessary, and post-capture assessment and manual labeling of corneal and IOL surface peaks on each axial scan is time consuming, adding up to 5 minutes of processing time to each set of scans. It is also difficult to use with dual-optic IOL systems because of multiple signal peaks. To our knowledge, our study is the first to compare ACD measurements by the Visante AS-OCT device in pseudophakic eyes with measurements by other instruments. The device is widely used and has a diverse range of applications in clinical and research settings. It has the benefit of capturing measurements in the fixating eye, and its axial resolution is 18 mm. It generates 2-D images of the anterior chamber; the images are constructed using an algorithm that overcomes image distortion induced by corneal optical transmission.18 The device does not contact the globe; however, angle-to-angle cross-sectional images can only be obtained in the horizontal plane without retracting the upper and lower lids. To obtain complete images away from the horizontal plane, lid retraction must be gently performed with fingertips or cotton buds, taking care not to distort the globe. In the present study, this was a little awkward and in some cases resulted in mild dry eye–type discomfort during measurements. According to the manufacturer, the ideal image is achieved when a vertical white line is visible on the anterior chamber image, indicating that the scan is oriented perpendicular to the apex of the cornea. This does not indicate that the image is centered along the visual or optical axis or the line of sight. Ideal images are not always possible with every scan. In quad scan mode (wherein 4 images are obtained at 45 degrees to each other), it is rare to simultaneously obtain 4 ideal images in a single capture. Once anterior segment images have been captured, ACD measurements are performed by manually aligning the anterior chamber tool during image analysis. The operator must place this tool in both iridocorneal angles and on the anterior lens surface. Thus, ACD measurements with the AS-OCT device are semisubjective. Baikoff et al.19 estimate that errors in alignment during ACD measurement amount to underestimation by approximately 20 mm, representing an error of 6% to 8%. This may have contributed to the shallower ACD measurements found with the AS-OCT device in the present study. Furthermore, in pseudophakic eyes with deep anterior chambers, the entire axial thickness of the IOL could not be determined because it did not fit within the frame of the image. This means that in some cases, the ACD may not be measured along the same axis under different accommodative conditions because lens thickness cannot be confirmed.24 Our study found that in pseudophakic eyes with an accommodating IOL, ACD measurements obtained by

the AS-OCT device were statistically significantly shallower than those obtained by the PCI device. This bias was more pronounced when ACD shallowed after administration of topical pilocarpine. The AS-OCT device’s bias toward underestimation of ACD measurements with reducing ACD resulted in overestimation of forward IOL movement and thus overestimation of the accommodative performance of the IOL. There are several potential explanations for the differences in ACD measurements between the 2 instruments. One of the most important is that each instrument samples different axial distances. The PCI device measures ACD parallel to the optical axis of the eye. The AS-OCT device measures ACD along a line running normal to the corneal apex under optimum image-capture conditions, even though optimum image capture is not always possible with the device. Another important source of error is the software algorithm the AS-OCT device uses to compensate for image distortion. Systematic errors in corneal thickness, ACD, and lens thickness measurements by the AS-OCT device have been identified by measuring standardized model eyes of known physical dimensions.25 Instrument distortion-correction factors have been proposed that are expected to improve successive generations of the device. After administration of topical pilocarpine, the AS-OCT device underestimated ACD by 63 mm and overestimated forward IOL movement by 36 mm. This difference was statistically significant, but relatively small in magnitude in the context of focus shift accommodating IOLs, in which 1000 mm of forward IOL movement would be required to produce 1.30 D of objective accommodation in an eye with an average AL.26 However, considering emerging strategies for the restoration of accommodation, such as lens refilling, that will better imitate the biometric changes observed in phakic accommodation, such differences between instruments could be relevant. Phakic accommodation is associated with steepening of the anterior and posterior lens surfaces, axial thickening, and an overall anterior movement of the crystalline lens. Baikoff et al.18 evaluated patients of various age with OCT and found that ACD decreased with accommodation by as much as 300 mm at 20 years, decreasing to 100 mm at 40 years. A similar magnitude of ACD change was observed with a laboratory PCI model in a group of patients between 23 years and 25 years of age. Near-point accommodation (8.70 D in this group) resulted in a mean decrease in ACD of 376 G 104 mm.27 Tsorbatzoglou et al.28 also used PCI to measure anterior segment changes with physiologic accommodative stimulation in a group of patients younger than 30 years. In the study, 3.00 D of

J CATARACT REFRACT SURG - VOL 36, MAY 2010

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

accommodative stimulation resulted in a mean ACD reduction of 80 G 60 mm. If a lens refilling system were able to restore 3.00 D of accommodation postoperatively by closely mimicking phakic accommodation, 80 mm of forward movement of the anterior lens surface, achieved by a combination of steepening of the anterior lens curvature and forwards lens movement, may be a realistic estimate of postoperative ACD change. Thus, the differences observed between the 2 instruments are likely to be of greater relevance when future generations of accommodating IOLs and lens-refilling strategies are assessed. Based on the results in the present study, measurements of ACD obtained by the PCI device and by the AS-OCT device are not interchangeable in pseudophakic eyes. Measurements with the AS-OCT device had a statistically significant bias toward a shallower ACD under all accommodative conditions, implying greater IOL movement. Estimates of axial IOL movement by the AS-OCT device should be interpreted with care. REFERENCES 1. Heatley CJ, Spalton DJ, Hancox J, Kumar A, Marshall J. Fellow eye comparison between the 1CU accommodative intraocular lens and the Acrysof MA30 monofocal intraocular lens. Am J Ophthalmol 2005; 140:207–213 2. Hancox J, Spalton D, Heatley C, Jayaram H, Marshall J. Objective measurement of intraocular lens movement and dioptric change with a focus shift accommodating intraocular lens. J Cataract Refract Surg 2006; 32:1098–1103 3. Wolffsohn JS, Naroo SA, Motwani NK, Shah S, Hunt OA, Mantry S, Sira M, Cunliffe IA, Benson MT. Subjective and objective performance of the Lenstec KH-3500 ‘‘accommodative’’ intraocular lens. Br J Ophthalmol 2006; 90:693–696. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1860198/pdf/ 693.pdf. Accessed January 17, 2010 4. Koeppl C, Findl O, Menapace R, Kriechbaum K, Wirtitsch M, Buehl W, Sacu S, Drexler W. Pilocarpine-induced shift of an accommodating intraocular lens: AT-45 Crystalens. J Cataract Refract Surg 2005; 31:1290–1297 5. Cumming JS, Colvard DM, Dell SJ, Doane J, Fine IH, Hoffman RS, Packer M, Slade SG. Clinical evaluation of the Crystalens AT-45 accommodating intraocular lens; results of the U.S. Food and Drug Administration clinical trial. J Cataract Refract Surg 2006; 32:812–825 6. Legeais JM, Werner L, Werner L, Abenhaim A, Renard G. Pseudoaccommodation: BioComFold versus a foldable silicone intraocular lens. J Cataract Refract Surg 1999; 25:262–267 7. Kriechbaum K, Findl O, Koeppl C, Menapace R, Drexler W. Stimulus-driven versus pilocarpine-induced biometric changes in pseudophakic eyes. Ophthalmology 2005; 112:453–459 8. McLeod SD, Vargas LG, Portney V, Ting A. Synchrony dualoptic accommodating intraocular lens. Part 1: optical and biomechanical principles and design considerations. J Cataract Refract Surg 2007; 33:37–46 9. Ossma IL, Galvis A, Vargas LG, Trager MJ, Vagefi MR, McLeod SD. Synchrony dual-optic accommodating intraocular lens. Part 2: pilot clinical evaluation. J Cataract Refract Surg 2007; 33:47–52

797

10. Nishi O, Nishi K, Nishi Y, Chang S. Capsular bag refilling using a new accommodating intraocular lens. J Cataract Refract Surg 2008; 34:302–309 11. Win-Hall DM, Ostrin LA, Kasthurirangan S, Glasser A. Objective accommodation measurement with the Grand Seiko and Hartinger coincidence refractometer. Optom Vis Sci 2007; 84:879–887 12. Wolffsohn JS, Hunt OA, Naroo S, Gilmartin B, Shah S, Cunliffe IA, Benson MT, Mantry S. Objective accommodative amplitude and dynamics with the 1CU accommodative intraocular lens. Invest Ophthalmol Vis Sci 2006; 47:1230–1235. Available at: http://www.iovs.org/cgi/reprint/47/3/1230. Accessed January 17, 2010 13. Win-Hall DM, Glasser A. Objective accommodation measurements in pseudophakic subjects using an autorefractor and an aberrometer. J Cataract Refract Surg 2009; 35:282–290 14. Drexler W, Baumgartner A, Findl O, Hitzenberger CK, Sattmann H, Fercher AF. Submicrometer precision biometry of the anterior segment of the human eye. Invest Ophthalmol Vis Sci 1997; 38:1304–1313. Available at: http://www.iovs.org/cgi/ reprint/38/7/1304. Accessed January 17, 2010 15. Kriechbaum K, Leydolt C, Findl O, Bolz M, Drexler W. Comparison of partial coherence interferometers: ACMaster versus laboratory prototype. J Refract Surg 2006; 22:811–816 16. Sacu S, Findl O, Buehl W, Kiss B, Gleiss A, Drexler W. Optical biometry of the anterior eye segment: interexaminer and intraexaminer reliability of ACMaster. J Cataract Refract Surg 2005; 31:2334–2339 17. Kriechbaum K, Findl O, Kiss B, Sacu S, Petternel V, Drexler W. Comparison of anterior chamber depth measurement methods in phakic and pseudophakic eyes. J Cataract Refract Surg 2003; 29:89–94 18. Baikoff G, Lutun E, Ferraz C, Wei J. Static and dynamic analysis of the anterior segment with optical coherence tomography. J Cataract Refract Surg 2004; 30:1843–1850 19. Baikoff G, Jodai HJ, Bourgeon G. Measurement of the internal diameter and depth of the anterior chamber: IOLMaster versus anterior chamber optical coherence tomographer. J Cataract Refract Surg 2005; 31:1722–1728 20. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307–310. Available at: http://www-users.york.ac.uk/ wmb55/meas/ba.pdf. Accessed January 17, 2010 21. Koranyi G, Lydahl E, Norrby S, Taube M. Anterior chamber depth measurement: A-scan versus optical methods. J Cataract Refract Surg 2002; 28:243–247 22. Nemeth G, Vajas A, Kolozsvari B, Berta A, Modis L Jr. Anterior chamber depth measurements in phakic and pseudophakic eyes: Pentacam versus ultrasound device. J Cataract Refract Surg 2006; 32:1331–1335 23. Marchini G, Pedrotti E, Modesti M, Visentin S, Tosi R. Anterior segment changes during accommodation in eyes with a monofocal intraocular lens: high-frequency ultrasound study. J Cataract Refract Surg 2008; 34:949–956 24. Schachar RA. Biometric changes in pseudophakic eyes [letter]. Ophthalmology 2005; 112:2241; reply by Findl O, Kriechbaum K 2241–2242 25. Dunne MCM, Davies LN, Wolffsohn JS. Accuracy of cornea and lens biometry using anterior segment optical coherence tomography. J Biomed Opt 2007; 12:064023 26. Nawa Y, Ueda T, Nakatsuka M, Tsuji H, Marutani H, Hara Y, Uozato H. Accommodation obtained per 1.0 mm forward movement of a posterior chamber intraocular lens. J Cataract Refract Surg 2003; 29:2069–2072

J CATARACT REFRACT SURG - VOL 36, MAY 2010

798

ANTERIOR CHAMBER DEPTH: PCI VERSUS AS-OCT

27. Koeppl C, Findl O, Kriechbaum K, Drexler W. Comparison of pilocarpine-induced and stimulus-driven accommodation in phakic eyes. Exp Eye Res 2005; 80:795–800 28. Tsorbatzoglou A, Ne´meth G, Sze´ll N, Biro´ Z, Berta A. Anterior segment changes with age and during accommodation measured with partial coherence interferometry. J Cataract Refract Surg 2007; 33:1597–1601

J CATARACT REFRACT SURG - VOL 36, MAY 2010

First Author: Georgia Cleary, MB BS, MRCOphth Department of Ophthalmology, St. Thomas’ Hospital, London, United Kingdom