Graefes Arch Clin Exp Ophthalmol (2009) 247:965–973 DOI 10.1007/s00417-009-1038-1
CATARACT
Comparison of contrast sensitivity, depth of field and ocular wavefront aberrations in eyes with an IOL with zero versus positive spherical aberration Jay S. Pepose & Mujtaba A. Qazi & Keith H. Edwards & Jeff P. Sanderson & Edwin J. Sarver
Received: 4 October 2008 / Revised: 7 January 2009 / Accepted: 12 January 2009 / Published online: 11 March 2009 # Springer-Verlag 2009
Abstract Purpose To compare the clinical performance of the zero spherical aberration (SA) SofPort LI61AO (AO, Bausch & Lomb) intraocular lens (IOL) to the AcrySof SA60AT (AT, Alcon), which has positive spherical aberration. Methods Patients underwent uneventful phacoemulsification with implantation of either an aspheric (AO, n=19) or spherical (AT, n = 20) IOL. Postoperatively, a 5 mm artificial pupil was positioned in trial frames with the cycloplegic refraction during monocular, mesopic contrast sensitivity (CSF) and low-contrast visual acuity (LCVA)
Supported in part by research grants from the Midwest Cornea Foundation, St. Louis, MO, USA and Bausch & Lomb, Rochester, NY, USA. Drs. Pepose, Qazi and Sanderson have no financial interest in the products described. Dr. Pepose has received research and travel support from Bausch & Lomb. Dr. Sarver has a financial interest in the VOL-CT Software used in this study. Mr. Edwards is an employee of Bausch & Lomb. Dr. Pepose is a paid consultant for Bausch & Lomb, Rochester, NY, USA J. S. Pepose (*) : M. A. Qazi : J. P. Sanderson Pepose Vision Institute, 1815 Clarkson Road, Chesterfield, MO 63107, USA e-mail:
[email protected] J. S. Pepose : M. A. Qazi Department of Ophthalmology & Visual Sciences, Washington University School of Medicine. St. Louis, Missouri, USA E. J. Sarver Sarver & Associates, Carbondale, Illinois, USA K. H. Edwards Bausch & Lomb, Rochester, New York, USA
testing with glare. Ocular and corneal wavefront error was determined at 5 mm diameters. Results Mean CSF scores were better at all frequencies tested for the AO than for the AT group, and achieved statistical significance at 1.5 cpd (p=0.038) and 6 cpd (p= 0.017). With glare, AO eyes read 30.9±5.0 low-contrast letters versus 25.2±6.8 for AT eyes (p=0.005) (mean ΔLogMAR = −0.10), while high-contrast acuity and refraction were similar. Eyes implanted with the SA60AT had 43% greater positive spherical aberration at a 5 mm wavefront diameter, with no significant difference in corneal SA between groups. A through-focus analysis demonstrated a similar depth of field, yet a comparatively higher visual Strehl ratio for the aspheric IOL at emmetropia (p=0.038). Conclusion Eyes with the SofPort Advance Optics neutral aberration IOL demonstrated less spherical aberration and better low-contrast acuity compared to eyes with a spherical IOL, without sacrificing tolerance to defocus. The aspheric IOL showed superior optical and clinical performance, which is most likely due to its surface design. Keywords Spherical aberration . Aspheric intraocular lens . Contrast sensitivity . Wavefront aberrometry
Introduction Spherical aberration (SA), a symmetrical higher-order optical aberration, has been identified as a key contributor to the deterioration of image quality, night myopia and photic complaints [1]. First corneal surface wavefront aberration analysis confirms that the prolate corneas of both young and cataract-age patients have, on average, positive spherical aberration [2–5]. In youth, the generally
966
negative SA of the crystalline lens largely neutralizes the average positive SA of the prolate cornea, resulting in an optimized retinal image [6–8]. In the aging eye, there is a loss of this cornea–lens balancing. As the growing crystalline lens matures, it develops positive SA which adds to, rather than offsets, the average positive corneal SA [8]. Similarly, conventional intraocular lenses (IOL), with their biconvex spherical design, typically add to corneal SA, resulting in suboptimal point spread and modulation transfer functions [9]. The driving force for the introduction of aspheric IOLs was an appreciation of the negative impact on image quality of an IOL with inherent positive spherical aberration [2]. Theoretical models have demonstrated the benefit of an aspheric IOL design in minimizing the induction of SA following cataract surgery [2, 10] and the impact of aspheric IOLs on whole eye SA has been demonstrated by wavefront analysis [11–33]. Aspheric IOLs have been designed with prolate anterior (e.g., Tecnis, Advanced Medical Optics, Santa Ana, CA, USA) or prolate posterior surfaces (AcrySof IQ, SN60WF, Alcon, Fort Worth, TX, USA), with varying degrees of negative SA designed to fully (Tecnis) [2] or partially (IQ) [34] offset the average corneal SA (+0.27 µm referenced to 555 nm monochromatic light at a 6 mm optical zone [2]). Another design strategy is the development of an aspheric IOL with prolate anterior and posterior surfaces that has neither positive nor negative SA (AO, SofPort AO [10], LI61AO, Bausch & Lomb, Rochester, NY, USA). This silicone IOL has equal optical power across the entire pseudophakos and, therefore, is neutral to the induction of positive or negative SA or other higher-order aberrations (HOA), even when decentered in relation to the visual axis [10]. We compared the clinical optical performance of the SofPort AO IOL with one of the most frequently implanted spherical, biconvex acrylic IOLs—the AcrySof SA60AT (AT, Alcon Laboratories, Fort Worth, TX, USA). Primary measures of outcome were mesopic contrast sensitivity (CSF) and low-contrast visual acuity with glare and wavefront aberrations. Since SA is more pronounced with larger pupil diameters [35], this variation in test conditions can have a profound influence on the outcomes of these examinations and on the optical quality of the retinal image. In this study, in an effort to minimize variations in pupil size that can affect performance on these optical and psychophysical tests, wavefront aberrometry was analyzed at fixed 5 mm diameter, and contrast sensitivity was similarly tested after dilation using a 5-mm aperture. As an alternative approach, others have fashioned black soft contact lenses of fixeddiameter apertures [25]. A multivariate analysis was performed to determine the relationship between whole-eye SA and mesopic low-contrast acuity, holding other higher-
Graefes Arch Clin Exp Ophthalmol (2009) 247:965–973
order aberrations constant. Finally, since some aspheric IOLs could negatively affect depth of field [36–39], a throughfocus analysis was performed to compare sensitivity of these lenses to a range of positive and negative defocus.
Patients and methods Entry and exclusion criteria Cataract patients with less than 2 diopters (D) of preoperative corneal astigmatism who had successfully undergone uncomplicated, sutureless, clear corneal incision phacoemulsification with implantation of either an aspheric, silicone IOL (AO, LI61A0, n=19 eyes) or with a biconvex, spherical acrylic IOL (AT, SA60AT, n=20 eyes) were examined for routine postoperative visits during a 2-month period. To avoid potential bias, patients meeting these criteria were sequentially enrolled. All patients entered into the study had retinal acuity meter tests of 20/20 preoperatively, were operated on by one surgeon (JSP) using a uniform technique, a 2.8-mm clear corneal temporal incision and standardized postoperative medication regimen. Pseudophakic eyes were enrolled for wavefront aberrometry and quality of vision testing if: (1) the bestcorrected visual acuity was 20/20 or better, (2) there was a well-centered posterior chamber intraocular lens viewed at the slit lamp, and (3) the posterior capsule was intact without evidence of significant PCO within the central 6 mm. Patients were excluded if there was a history of diabetes, dry eye, corneal dystrophy, corneal opacity or edema, corneal surface irregularity (e.g., nodules, ectasia, basement membrane dystrophy), glaucoma or other optic nerve disease, cystoid macular edema, retinal detachment, macular hole or pseudohole, age-related macular degeneration or other maculopathy. This research was approved by an Institution Review Board, and all patients were entered into the study after giving an informed consent. Diagnostic testing All patients underwent both wavefront and contrast sensitivity testing. Prior to dilation with one drop each of mydriacyl 0.5% (Bausch & Lomb) and phenylephrine 2.5% (Bausch & Lomb), uncorrected vision (UCVA) on an Early Treatment of Diabetic Retinopathy Study (ETDRS) chart under standard lighting conditions, manifest refraction with best-corrected visual acuity (BCVA), scotopic pupillometry (Colvard; Oasis Medical, Glendora, CA, USA) and corneal videokeratography (Orbscan, Bausch & Lomb) were obtained. Following dilation to at least 5 mm, a cycloplegic refraction was repeated. A 5 mm artificial pupil was positioned and centered in trial frames by the patient, with
Graefes Arch Clin Exp Ophthalmol (2009) 247:965–973
a refined cycloplegic refraction in place during monocular, mesopic contrast sensitivity (CSF) with glare and mesopic low-contrast acuity (LCVA) testing with glare on the Optec 6500 (Vision Sciences Research, San Ramone, CA, USA). The patients were tested at 3 cd/m [2] following 10 minutes of dark adaptation, with a glare source of 3 lux. The effect of spherical aberration on LCVA was investigated by multiple stepwise backward regression analysis, which was performed with the differences in LCVA as the dependent variable, and with coma and other higher-order aberrations as independent variables. The measurements were taken with the observer masked to the IOL type. Ocular wavefront error was measured with Hartmann– Shack aberrometry (COAS, WaveFront Sciences, Albuquerque, NM, USA; and Zywave, Bausch & Lomb), and are reported following the ANSI Z80.28 standard for specifying ocular aberrations of the eye. Wavefront measurements at 780 nm were referenced to 555 nm, and exported from the aberrometer as Zernike coefficients fitted up to the fifth order for a maximum pupil size. Through scaling of the coefficients, each Zernike coefficient was determined at a 5 mm pupil size. Text files from the Orbscan (txt format of true anterior elevation data using the Recorder tool) were imported into custom software (VOLCT version 6.89, Sarver and Associates, Carbondale, IL, USA) to determine the corneal wavefront. Mean levels of individual ocular and corneal Zernike coefficients at 5 mm wavefront diameters were obtained for the two study groups and statistically analyzed. Using VOL-CT software, a through focus analysis ranging from +1.00D to -1.00D was performed plotting the visual Strehl (VSOTF [40], computed in the spatial frequency domain by weighting the ocular transfer function by the neural contrast sensitivity function) versus defocus for the aspheric and spherical IOL groups. To compare the depth of field, the visual Strehl ratio was used as an image-quality metric. The astigmatism was removed, and the visual Strehl ratio was calculated as a function of defocus from the Zernike coefficients using Fourier optics. The depth of field was then defined as the defocus interval at which the visual Strehl ratio was above or equal to 80% of the maximal value.
Results IOL designs and patient demographics Some of the differentiating design characteristics of the AT and AO IOLs are compared in Table 1. There were 19 eyes of 16 patients and 20 eyes of 15 patients consecutively enrolled into the AO and AT groups respectively. During enrollment, five patients implanted with each of the two IOL types were excluded because of retinal conditions that precluded 20/20
967 Table 1 IOL characteristics
Optic Lens Design Haptic material Optic material Overall length Haptic angulation Refractive index Optic diameter
AcrySof SA60AT
SofPort AO
Monofocal 1-piece Biconvex spherical surfaces Hydrophobic acrylic Hydrophobic acrylic 13 mm 0 degrees 1.55 6 mm
Monofocal 3-piece Conic aspheric anterior and posterior surfaces PMMA Silicone 13 mm 5 degrees 1.43 6 mm
best-spectacle corrected vision. There was no statistical difference between the two sets of postoperative eyes with respect to age, postoperative UCVA, BCVA, spherical equivalent, astigmatism, or scotopic pupil size (Table 2). Mean power of the implanted IOLs was not statistically different between the two groups (p=0.43). The median IOL power was 19.0D (range 11.0 to 27.5) and 20.0D (range 6.0D to 31D) for the AO and AT eyes respectively. Evaluation of first surface corneal wavefront A Zernike decomposition of the corneal wavefront was determined for eyes in each group at a 5 and 6 mm optical zone using Orbscan anterior elevation data (Table 3). There were no statistical differences in second, third or fourth order corneal aberrations between the two groups. Mean corneal fourth order SA at 6 mm was +0.30 and +0.27 for the AO and AT groups respectively (p=0.34), consistent with previously published average values [1, 4]. Total-eye wavefront aberrations With the COAS aberrometer at a 5 mm wavefront diameter, fourth order SA (Z40) was selectively increased in a statistically significant manner by 43% in spherical IOL eyes (mean 0.50AT, 0.35 µm AO, p=0.03). Total aberrations were 1.49±0.66 µm and 1.45±0.63 µm (p=0.89) and higher-order aberrations were 0.31±0.10 µm and 0.35± 0.09 µm (p=0.27) for AT and AO eyes respectively at a 5 mm wavefront diameter. There was no statistical difference in second or third order aberrations between the SA60AT and LI61AO groups (p>0.32, Table 4). The M, J0 and J45 components closely matched the cycloplegic refraction. Contrast sensitivity and low-contrast acuity CSF and LCVA were measured monocularly after dilation, with best spectacle correction in trial frames in mesopic
968
Graefes Arch Clin Exp Ophthalmol (2009) 247:965–973
Table 2 Demographics and postoperative features of aspheric intraocular lens (LI61AO) and biconvex intraocular lens (SA60AT) groups
P= probability value using t-test
Eyes/patients Female Age (years) Months postoperative Uncorrected visual acuity (20/) Best-corrected visual acuity (20/) Spherical equivalent (D) Cylinder (D) Mean Orbscan average SimK (D) Orbscan average cylinder (D) Mesopic pupil (mm) IOL power (D)
lighting conditions. A 5 mm artificial pupil was also placed into the trial frame to control for pupil size. Mean mesopic contrast sensitivity scores with glare were better for all frequencies tested in the AO group. These differences were statistically significant at 1.5 and 6 cycles per degree (cpd) (Table 5), and the same trend was seen at 3 cpd, although statistical significance was not demonstrated. Eyes with the aspheric IOL had better mean low contrast acuity (with glare) than their non-aspheric counterparts, reading an average of 30.9 versus 25.2 letters (p=0.0047)—a difference of −0.1 LogMAR.
There is an inverse correlation between the total ocular spherical aberration and mesopic low contrast visual acuity with glare (LCVA) for the aspheric IOL (R2 =0.376) but not for the spherical IOL. This indicates that for the IOL with no SA, 37% of the variation in low contrast acuity in Table 3 Corneal wavefront aberrations at a 5 mm diameter
Z20 Z3−3 Z3−1 Z31 Z33 Z40 HOA
SA60AT 5mm
P
Mean (µm)
SD
Mean (µm)
SD
0.273 −0.081 −0.089 0.101 −0.041 0.142 0.450
0.502 0.075 0.174 0.211 0.123 0.067 0.168
0.324 −0.053 −0.006 0.034 −0.047 0.132 0.426
0.497 0.189 0.218 0.155 0.167 0.060 0.134
SA60AT
P
19/16 6 64.8±7.4 4.6±2.7 42.5±38.3 17.5±2.5 −0.39±0.86 −0.66±0.62 44.76±1.93 1.16±0.72 4.70±1.20 17.71±4.39
20/15 7 68.4±10.5 6.9±4.2 34.4±40.3 18.5±2.5 −0.59±0.73 −0.46±0.32 44.68±1.31 0.93±0.48 4.67±1.15 19.19±6.64
0.23 0.05 0.52 0.22 0.44 0.23 0.89 0.25 0.97 0.43
the eyes with the aspheric IOL can be attributed to whole eye SA (Fig 1). A multivariate analysis holding coma and other higher-order aberration constants reveals that these other variables were not trivial explanatory factors for this statistically significant negative correlation between spherical aberration and mesopic LCVA in the aspheric IOL group. In contrast, for eyes with considerable positive SA introduced by the spherical IOL compounding the inherent positive corneal SA, there is no apparent correlation between LCVA and whole eye SA (R2 =0.002). Relationship between IOL power and whole-eye spherical aberration
Impact of whole-eye spherical aberration of mesopic low-contrast visual acuity
LI61AO 5mm
LI61AO
0.751 0.545 0.191 0.259 0.893 0.641 0.617
Z20 =defocus, Z3−3 = vertical trefoil, Z3−1 = vertical coma, Z31 = horizontal coma, Z33 = horizontal trefoil, Z40 =4th order spherical aberration, HOA= total higher-order aberrations, SD = standard deviation, P= probability value using 2-sided Student’s t-test
Figure 2 shows the relationship between whole-eye SA at 5 mm and IOL power for the spherical and aspheric IOL groups. Not surprisingly, the aspheric IOL of neutral SA shows no significant correlation between whole eye SA and IOL power. In contrast, there is a positive statistically significant (p
