opo 12133

8/17/2019 Opo 12133

1/12

Optical quality and visual performance with customised softcontact lenses for keratoconusAmit Jinabhai 1 , Clare O’Donnell 1,2,3 , Cindy Tromans 1,4 and Hema Radhakrishnan 1

1 Faculty of Life Sciences, The University of Manchester, Manchester, UK 2 Optegra Eye Sciences, Optegra Eye Hospital, Manchester, UK 3 School of Lifeand Health Sciences, Aston University, Birmingham, UK and 4 Manchester Academic Health Science Centre, Central Manchester University HospitalsNHS Foundation Trust, Manchester, UK

Citation information: Jinabhai A, O’Donnell C, Tromans C, Radhakrishnan H. Optical quality and visual performance with customised soft contact

lenses for keratoconus. Ophthalmic Physiol Opt 2014. doi: 10.1111/opo.12133

Keywords: aberration-controlling contactlenses, coma aberrations, higher-orderaberrations, keratoconus, rigid gas-permeablecontact lenses, toric soft contact lenses

Correspondence : Hema RadhakrishnanE-mail address: [email protected]

Received: 16 July 2013; Accepted: 18 March2014

Abstract

Purpose: This study investigated how aberration-controlling, customised softcontact lenses corrected higher-order ocular aberrations and visual performance

in keratoconic patients compared to other forms of refractive correction (specta-cles and rigid gas-permeable lenses). Methods: Twenty-two patients (16 rigid gas-permeable contact lens wearers andsix spectacle wearers) were tted with standard toric soft lenses and customisedlenses (designed to correct 3rd-order coma aberrations). In the rigid gas-perme-able lens-wearing patients, ocular aberrations were measured without lenses, withthe patient’s habitual lenses and with the study lenses (Hartmann-Shack aberrom-etry). In the spectacle-wearing patients, ocular aberrations were measured bothwith and without the study lenses. LogMAR visual acuity (high-contrast and low-contrast) was evaluated with the patient wearing their habitual correction (of either spectacles or rigid gas-permeable contact lenses) and with the study lenses.Results: In the contact lens wearers, the habitual rigid gas-permeable lenses andcustomised lenses provided signicant reductions in 3rd-order coma root-mean-square (RMS) error, 3rd-order RMS and higher-order RMS error ( p ≤ 0.004). Inthe spectacle wearers, the standard toric lenses and customised lenses signicantly reduced 3rd-order RMS and higher-order RMS errors ( p ≤ 0.005). The spectaclewearers showed no signicant differences in visual performance measuredbetween their habitual spectacles and the study lenses. However, in the contactlens wearers, the habitual rigid gas-permeable lenses and standard toric lensesprovided signicantly better high-contrast acuities compared to the customisedlenses ( p ≤ 0.006).Conclusions: The customised lenses provided substantial reductions in ocularaberrations in these keratoconic patients; however, the poor visual performancesachieved with these lenses are most likely to be due to small, on-eye lens decentra-tions.

Introduction

Optimal visual correction in keratoconus varies dependingon disease severity.1 Patients with mild keratoconus may achieve satisfactory visual acuity using spectacles. However,as the cornea becomes progressively more irregular, specta-cle-correction becomes increasingly unsuccessful, due to

large magnitudes of corneal astigmatism, 2 corneal apicalscarring 3 and ocular higher-order aberrations, in particularvertical coma. 4,5 Although alternative modalities (includinghybrid 6,7 and mini-scleral 8,9 lenses) have been used to man-age patients with keratoconus, rigid gas-permeable (RGP)lenses are currently considered to be the ‘gold standard’ fornon-surgical visual rehabilitation of patients with moderate

© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists 1

Ophthalmic & Physiological Optics ISSN 0275-5408

8/17/2019 Opo 12133

2/12

to severe disease.10 – 12 When applied to the eye, RGP lenses

mask most of the induced anterior corneal surface aberra-tions by replacing the irregular keratoconic corneal surfacewith the smooth, regular refractive surfaces of the RGP lensand a liquid tear-lens. 13,14 However, several studies haveshown that residual higher-order aberrations persist even

with RGP lenses on-eye,4,15 – 17

which are typically attributedto the posterior corneal surface. 15,18,19

Using adaptive optics, previous studies have demon-strated that correction of higher-order aberrations resultsin a signicant improvement in both high- and low-con-trast acuity in keratoconic patients. 20,21 Arguably, suchimprovements in visual performance would enhance every-day visual tasks, such as night-time driving. To optimiseoptical quality in this patient demographic, several reportshave investigated the use of customised, aberration-con-trolling soft contact lenses (ACCLs) to reduce ocular aber-rations. 22

– 28 However, some studies did not measure visualperformance, whilst others only used the metric of higher-order root-mean-square error (RMS) to evaluate changesin optical quality. Customised ACCLs may offer certainadvantages over RGP lenses, such as improved comfort(particularly in cases of RGP lens intolerance) and longerwearing times.17 Furthermore, young ametropic adults arelikely to achieve a better quality-of-life whilst wearing con-tact lenses rather than spectacles. 29 This study investigateshow standard, lathe-cut toric soft lenses (TCLs), specically designed for irregular corneas, and customised ACCLs cor-rect ocular higher-order aberrations and impact on visualperformance in keratoconus patients, compared to habitualRGP lenses or spectacles. The ACCLs were designed to

either fully (100%) or partially (50%) correct 3rd-ordercoma (vertical and horizontal) aberrations.

In addition to higher-order RMS error, this study exploreschanges in individual Zernike coma coefcient terms, comaRMS error, trefoil RMS error and spherical aberration. Thisstudy uses the same natural pupil size to compare aberrationsbetween subject groups without the use of a mydriatic agent,as pharmacological pupil dilation can impact on aberrationsby causing a shift in the location of the pupil centre. 30,31

Methods

Twenty-two keratoconic patients participated in this study;subjects were recruited from the Optometry Clinics at TheUniversity of Manchester and Manchester Royal Eye Hospi-tal. To be eligible for enrolment, patients had to meet twoor more of the inclusion criteria which comprised a‘scissors’ retinoscopic reex, distorted keratometric miresand slit lamp biomicroscopic signs of keratoconus – e.g.corneal protrusion, Vogt’s striae, apical scarring and Flei-scher’s ring. Patients with any other pathology or history of ocular surgery were excluded. All experimental data were

collected from one eye of each patient; the eye selectedshowed more advanced disease compared to the contralat-eral eye. Sixteen participants were successful habitual RGPlens wearers; the remaining six patients habitually worespectacles for vision correction. Successful wearers weredened as patients who experienced no more than slight

lens awareness and/or occasional foreign body sensationwhilst wearing their lenses, and regularly wore lenses for aminimum of eight hours a day. 32 Five females and 17 maleswere recruited; the mean age was 34 years (range 19 – 55 years).The investigation followed the tenets of the Declaration of Helsinki. Informed consent was obtained after explanationof the nature and possible consequences of the study. Thisinvestigation was approved by the Northwest 11 ResearchEthics Committee, of the National Research Ethics Service(NRES). Figure 1 displays a owchart summarising thesequence of investigative steps for either patient group.

Visual performanceHigh-contrast – 95% (HCVA) and low-contrast – 15%visual acuities (LCVA) were measured using Bailey-LovielogMAR charts (National Vision Research Institute; http://www.nvri.org.au/pages/-products-logmar-charts-and-more-.html) positioned at three metres – illuminated to approxi-mately 750 lux without causing glare. All measurementswere scored to the letter and adjusted to a six-metre testingdistance. The patient’s pupil size, whilst viewing the dis-tance HCVA chart, was also recorded (pupil gauge).

The SKILL (the Smith-Kettlewell Institute Low Lumi-nance) card (http://www.ski.org/) was used to evaluate

near acuity in both high-contrast and simulated, reduced-luminance conditions. 33 Acuities were recorded using thehigh-contrast and reduced-luminance characters at 40 cm,with an appropriate reading addition where necessary.Acuities were measured using the ‘letter-by-letter’ method;the nal SKILL card ‘score’ was calculated as the differencein the number of letters correctly identied between thetwo charts. 33 All visual performance measurements weremade in a consulting room with a constant illumination of approximately 310 lux.

Ocular aberrations

Ocular aberrations were evaluated using the IRX-3,Hartmann-Shack aberrometer (ImagineEyes; http://www.imagine-eyes.com/content/view/38/94/). This device uses a32-by-32 lenslet array and near infrared light (780 nm).Aberrations were recorded under monocular conditionswith the room lights switched off. Aberrations werecomputed up to the 6th Zernike order, for a 4-mm pupildiameter, using the device’s software (v.1.2; ImagineEyes;http://www.imagine-eyes.com/content/view/25/50/). A 4-mm

© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists2

Customised soft contact lenses for keratoconus A Jinabhai et al.

8/17/2019 Opo 12133

3/12

8/17/2019 Opo 12133

4/12

between 45 and 52 D were graded as ‘moderate’ keratoc-onus and readings greater than 52 D were classed as ‘severe’keratoconus. 36

After collecting baseline data, the 16 contact lens wearerswere instructed to leave their RGP lenses out for 7 days toact as a ‘wash-out’ period prior to tting the study lenses. 37

All 16 patients had spectacles, which had been prescribedwithin the previous 18 months and were asked to use theseuntil the trial lens tting appointment exactly one week later.

Soft trial lens tting assessment and over-refraction

Using each patient’s corneal topography as a guide, a seriesof different Plano trial lenses were applied and evaluatedon-eye to ascertain the best tting lens characteristics withregard to lens centration (horizontally and vertically),movement (digital push-up test, with blink in the primary position and upon up-gaze, and both horizontal and verti-cal lag) and corneal coverage, using established criteria. 38

Lens rotations in the primary position and upon blinkingwere evaluated by observing the position of the lens orien-tation marker. After a settling period of approximately 30 min, a monocular over-refraction was performedthrough the best-tting trial lens, at 6 m (Snellen chart;http://www.precision-vision.com/index.cfm/product/272_114/6-meter-20-ft-traditional-snellen-eye-chart.cfm; lumi-nance = 120 cdm 2 ), to obtain the subjective sphero-cylin-drical refraction that maintained optimal visual acuity.Retinoscopy with the trial contact lens was carried out toascertain a starting point for the over-refraction. This resultwas rened using established techniques, tailored to kerato-

conic patients with poorer acuity,17,39,40

to achieve themaximum ‘plus’ sphero-cylindrical power which allowedpatients to read their threshold visual acuity. Both logMAR HCVA and LCVA were then recorded (see visual perfor-mance section). The best-tting trial lens’ back optic zoneradius, peripheral design, laser marker orientation (indegrees) and subjective over-refraction (compensated forvertex distance) were used to order a TCL for each patient.The manufacturer appropriately adjusted the TCL’s nalcylinder axis to compensate for any small, on-eye lens rota-tions observed during tting (of up to 5 degrees).

On completion of the trial lens tting, the 16 RGP lens-wearing patients were asked to resume RGP lens wear asnormal whilst the TCLs were ordered. However, beforetheir collection appointment, the lens-wearing patients wereinstructed to leave their habitual RGP lenses out overnight,and to attend their appointment wearing their spectacles.

Standard toric soft lens (TCL) collection appointment

On average, the TCL collection appointment took placetwo weeks after the trial lens tting. Five minutes after

applying the TCL, a slit-lamp biomicroscopic evaluationwas performed to ensure that each lens was positionedacceptably in terms of lens centration, coverage and toricmarker alignment. Each TCL tted was designed with aprism-ballasted, toric front surface design, a spherical back surface and a total diameter of 14.50 mm. All TCLs

were manufactured using a Filcon II-based material (watercontent 77%, Dk = 53 and centre thickness of 0.40 mm).Thirty minutes after application, a slit-lamp examination

was performed to reassess the lens’ movement and posi-tion, to ensure that these had not altered substantially overthis period. A spherical over-refraction was then performedat six metres (Snellen chart). Where required, the over-refraction was kept in place to record logMAR HCVA,LCVA, and SKILL card scores.

The TCLs tted in this investigation were used todesign the customised ACCLs; they also served as a con-trol. Aberrations, with the TCLs, were recorded using theIRX-3 device; any residual 3rd-order coma aberrationswere used to manufacture two additional ACCLs for eachpatient. The ACCLs were made from the same materialas the TCLs, and were designed with non-rotationally symmetric surfaces, containing the previously establishedsphero-cylindrical correction, and either a 100% (referredto hereafter as 100% lenses) or a 50% correction(referred to hereafter as 50% lenses) for the residualcoma aberrations. On average, residual vertical and hori-zontal coma aberrations in either group were negative insign; therefore, the ACCLs were each designed with therequired magnitudes of positive coma (on the front sur-face) to reduce residual aberrations. Both the 100% and

50% ACCLs were veried against the desired aberrationdesigns using Hartmann-Shack aberrometry (ClearWaveaberrometer; http://www.wfsci.com/clearwave.htm). 41,42

Before the customised lens collection appointment, thelens-wearing patients were again instructed to leave theirhabitual lenses out overnight, and to attend their appoint-ment wearing their spectacles.

Customised lens collection appointment

To minimise bias, the order in which the two ACCLs weretted was randomised; the study patients were also maskedfrom knowing the orders. To ensure clinically acceptable lenscentration, movement, coverage and orientation, a slit-lampevaluation was performed at 5 and 30 min after applying therst ACCL. A spherical over-refraction was then performedat 6 m (Snellen chart). Where required, the over-refractionwas kept in place to record logMAR HCVA, LCVA andSKILL card scores. Aberrations were then evaluated with therst ACCL, this lens was then removed and the secondACCL applied. The same procedures and measurements, asoutlined above, were then repeated for the second ACCL.



8/17/2019 Opo 12133

5/12

Data analysis

Higher-order RMS aberrations were calculated as theroot-mean-square of all Zernike coefcients from the3rd- to the 6th-order, inclusive. Normality was evaluatedusing the Shapiro-Wilk’s test, for a critical value of 0.05

(v.16.0; SPSS, http://www-01.ibm.com/software/analytics/spss/products/statistics/). Normally-distributed data wereanalysed with 1-way repeated-measures analysis of vari-ance (RM-ANOVA) – critical value of 0.05. Non-normally-distributed data were evaluated with the Wilcoxonsigned-ranks test and the Friedman test (critical value of 0.05 for both).

In the RGP lens wearers, post-hoc analyses for normally-distributed data were conducted using a Bonferroni adjust-ment for 10 multiple comparisons (critical value of 0.005).For non-normally-distributed data, post-hoc analyses wereconducted using the Wilcoxon signed-ranks test (criticalvalue of 0.005). In the spectacle wearers, post-hoc analysesfor the normally-distributed aberrometry data were con-ducted using a Bonferroni adjustment for six multiple com-parisons (critical value of 0.0083). For correlation analyses,normally-distributed data were analysed using two-tailedPearson’s correlation ( RP ), whereas non-normally-distrib-uted data were evaluated using two-tailed Spearman’scorrelation ( RS); critical value of 0.05 for both.

Results

The central radii, central corneal thicknesses, diseaseseverity grades,36 corneal slit-lamp appearances and RGP

lens tting grades35

(at the initial examination) are pre-sented in the Table S1, Supporting Information. All 16RGP lens wearers were tted with the TCLs. CustomisedACCLs could not be manufactured for patient 3, as themean residual vertical and horizontal coma aberrationswith the TCLs were extremely small (+ 0.008 and+ 0.02 l m, respectively). A few measurements with theACCLs could not be obtained, as some patients wereunable to attend all the scheduled visits. Fourteen of the16 RGP lens wearers were tted with the 100% lenses(patients 1, 2, 4, and 6 – 16), while 13 patients were ttedwith the 50% lenses (patients 1, 2, 4, 6 – 11 and 13 – 16).Although ocular aberrations could not be recorded witheither ACCL type for patient 4, visual performance mea-surements were possible. Unobtainable results were trea-ted as ‘missing’ data. Consequently, aberrometry datameasured across the four contact lens types (RGP, TCL,100% and 50%) could only be compared for 12 (of the16) RGP lens wearers, whereas visual performance datawere analysed for 13 (out of 16) subjects.

All visual acuity and aberration measurements wereobtained from the six spectacle-wearers.

Ocular aberrations

RGP lens-wearersIn the RGP lens wearers, 2nd-order cylinder RMS (RMS of C2,+ 2 and C2, 2); defocus; 3rd-order coma RMS and 4th-order spherical aberration each showed signicant

differences between the ve measurement conditions( p ≤ 0.006; RM-ANOVA).The RGP lenses signicantly reduced 2nd-order cylinder

RMS and defocus ( p = 0.003, Figure 2). In contrast to theRGP lenses, all three study lenses induced 2nd-order cylin-der RMS ( p ≤ 0.001); however, only the 100% lensesinduced signicantly more cylinder RMS error than whenuncorrected ( p = 0.004).

The RGP lenses and both customised lenses signicantly reduced coma RMS error ( p ≤ 0.001; Figure 3); however, theTCLs induced signicant negative spherical aberration( p = 0.003). Compared to the RGP lenses, both customisedlenses also induced negative spherical aberration ( p ≤ 0.002).

Friedman analyses revealed signicant differencesbetween the ve measurement conditions for trefoil RMS,3rd-order RMS, 4th-order cylinder RMS and higher-orderRMS error ( p ≤ 0.033). The RGP lenses and TCLs signi-cantly reduced trefoil RMS error ( p ≤ 0.002, Figure 3).The 100%, 50% and RGP lenses also signicantly reduced3rd-order RMS error ( p ≤ 0.004, Figure 3), whereas only the 50% lenses and RGP lenses signicantly reduced4th-order cylinder RMS error ( p ≤ 0.003). All four lenstypes signicantly reduced higher-order RMS error( p ≤ 0.004).

Friedman analyses showed a signicant difference in verti-

cal coma between the ve measurement conditions ( v 2 =40.3; p < 0.0001). Post-hoc analyses (Wilcoxon signed-ranks test) showed signicant differences in vertical coma,from 0.93 0.34 l m [ 1 S.D.] when uncorrected, to+ 0.18 0.39 l m with the 100% lenses ( Z = 3.1, p = 0.002); to 0.17 0.30 l m with the 50% lenses( Z = 3.1, p = 0.002) and to + 0.39 0.14 l m with theRGP lenses ( Z = 3.1, p = 0.002). In contrast, the TCLs didnot signicantly reduce vertical coma ( 0.66 0.43 l m; p = 0.04). Both the 100% and RGP lenses produced a posi-tive shift in vertical coma; although the RGP lenses showedhigher residual aberration values, the differences were notstatistically signicant ( p = 0.14).

Freidman analyses revealed no signicant differences inhorizontal coma between the ve measurement conditions( p = 0.10).

Spectacle wearersIn the spectacle wearers, all the calculated higher-orderaberration metrics showed signicant differences betweenthe four measurement conditions ( p ≤ 0.033; RM-ANOVA).In contrast, there were no signicant differences in either


A Jinabhai et al. Customised soft contact lenses for keratoconus

8/17/2019 Opo 12133

6/12

Figure 2. Correction of 2nd-order cylinder RMS and defocus aberrations measured without lenses, with the patients’ habitual RGP lenses, standardtoric soft lenses (TCLs) and both the 100% and 50% lenses, in the contact lens wearers (left-hand gure; n = 12). Correction of 2nd-order cylinderRMS and defocus aberrations measured without lenses, with the TCLs and both the 100% and 50% lenses, in the spectacle wearers (right-hand gure;n = 6). Error bars represent 1 standard deviation. RMS = root-mean-square error; RM-ANOVA = repeated-measures analysis of variance.

Figure 3. Comparing the correction of higher-order aberrations measured without lenses, with the patients’ habitual RGP lenses, standard toric softlenses (TCLs) and both the 100% and 50% lenses in the contact lens-wearing group. Error bars represent 1 standard deviation. RMS = root-mean-square error; RM-ANOVA = repeated-measures analysis of variance; WSRT = Wilcoxon signed-ranks tests.



8/17/2019 Opo 12133

7/12

defocus or 2nd-order cylinder RMS error ( p ≥ 0.32;Figure 2).

The 100% lenses signicantly reduced 3rd-order comaRMS error ( p = 0.005, Figure 4), whereas no signicantdifferences were found with either the TCLs or 50% lenses.Both 3rd-order RMS and higher-order RMS error were sig-nicantly reduced by each of the three study lenses( p ≤ 0.005).

RM-ANOVA showed a signicant difference in vertical comabetween the four measurement conditions ( F 3,15 = 9.6, p = 0.001). The 100% lenses induced a signicant positive

shift in vertical coma, from 0.54 0.29 when uncor-rected, to + 0.21 0.24 l m ( p = 0.007), whereas both the50% lenses (+ 0.01 0.30 l m) and TCLs ( 0.31 0.21 l m) induced non-signicant positive shifts ( p ≥ 0.01).

Freidman analyses showed no signicant differences inhorizontal coma between the four measurement conditions( p = 0.20).

Visual performance

Figure 5 displays the visual acuity results in the contact lensgroup, where signicant differences in logMAR HCVA,LCVA and SKILL card scores were recorded between thefour lens types (F 3,36 ≥ 7.9, p ≤ 0.0003: RM-ANOVA). The RGPlenses provided signicantly better HCVA than the 100%lenses ( p = 0.002). Similarly, the TCLs provided signicantly better acuities than both the 100% and 50% lenses( p ≤ 0.006). The LCVA results also showed that RGP lensesperformed signicantly better than the 100% lenses( p = 0.001). In addition, the SKILL card scores were foundto be signicantly lower with RGP lenses compared to eitherthe TCLs or 100% lenses ( p ≤ 0.001).

In the non-lens wearers, the patient’s habitual spectaclesgave the best visual performances, where meanHCVA = + 0.20 0.19 logMAR; mean LCVA = + 0.51 0.19 logMAR and mean SKILL card score = 29 6 letters.Similar values were also observed with the three study lenses. There were no signicant differences in logMAR HCVA, LCVA and SKILL card scores between the fourmeasurement conditions ( p ≥ 0.06; see Figure S1, Support-ing Information for further details).

The mean pupil size recorded whilst measuring logMAR HCVA was 4.1 0.4 mm [ 1 S.D.] n = 22. Over-

refractions used to maximise visual performance with eitherthe RGP lenses or TCLs ranged from 0.50 to + 0.75 D.However, the over-refractions recorded with the customisedlenses showed larger variations (100% lenses: between 1.25and + 1.00 D and 50% lenses: between 0.75 and + 0.75 D).The larger variation with the 100% lenses might be due tocompensation for the slightly higher residual RMS aberra-tions measured with these lenses on-eye.

Correlations between visual performances and ocular aberra-tionsBoth HCVA and LCVA, measured with each form of opti-cal correction (spectacles/RGP lenses, TCLs, 100% lensesand 50% lenses), were plotted against the ocular aberrationterms recorded across both study groups. All statistically signicant relationships found between these parametersare summarised in the Table S2, Supporting Information.

Discussion

In the contact lens wearers, the patient’s habitual RGPlenses provided signicantly lower SKILL card scores

Figure 4. Comparing the correction of higher-order aberrations measured without lenses, with the standard toric soft lenses (TCLs) and both the100% and 50% lenses in the spectacle-wearing group. Error bars represent 1 standard deviation. RMS = root-mean-square error; RM-ANOVA =

repeated-measures analysis of variance.



8/17/2019 Opo 12133

8/12

compared to the TCLs. Although HCVA and LCVA scoreswere also noticeably better with the RGP lenses, the differ-ences were not statistically signicant. Similar improve-ments in visual performances, over TCLs, have also beenreported in previous studies of keratoconic patients. 17,26

Such improvements are attributed to a superior correction

of corneal astigmatism with RGP lenses. 17,43 In compari-son, larger magnitudes of residual astigmatism would per-sist with TCLs, as hydrogel lenses exert less pressure ontothe irregular cone apex, thereby limiting the masking of astigmatism. 44

– 46 The high levels of astigmatism measuredwith the TCLs on-eye might also be attributable to small

degrees of on-eye lens rotation. Such residual cylindricalerrors appear to impact on visual performance, as a signi-cant correlation was found between logMAR LCVA and2nd-order cylinder RMS error measured with the TCLs(Table S2, Supporting Information).

Another factor inuencing visual performance with RGPlenses is the lens’ tting, as previous studies have demon-strated that atter-tting lenses provide better visual acuity in keratoconic patients than steeper-tting lenses. 14,46,47

Fifteen of our 16 RGP lens wearers presented with anapical-bearing RGP lens tting.

In the non-lens wearers, mean HCVA scores were similarbetween spectacles and TCLs, whilst LCVA and SKILL cardscores were slightly better with spectacles than with TCLs.These results are also likely to be due to larger magnitudesof astigmatism persisting with the TCLs. In agreement withJinabhai et al.,48 the high magnitudes of 2nd-order cylinderRMS error recorded with the TCLs (in both patient groups)showed a large degree of variability, which is typically attributed to spot-imaging errors at the wavefront sensor.

Our ndings on changes in HCVA and LCVA with thetwo ACCLs conict with previous investigations. L opez-Gilet al.25 found that customised lenses provided signicantly better HCVA compared to spectacles, whilst Sabesanet al.23 reported that customised lenses provided improved

LCVA (logMAR) compared to conventional TCLs andhabitual RGP lenses. Marsack et al.27 found that bothHCVA and LCVA improved with customised lenses com-pared to conventional TCLs; however, unlike Sabesanet al.23 , Marsack et al.27 found that customised lensesimproved HCVA more than LCVA. Like Sabesan et al.23 ,Katsoulos et al.24 reported larger improvements in LCVAthan HCVA whilst comparing customised lenses to specta-cle refraction. Dissimilarities between our results and thoseof previous reports might be due to differences in contactlens materials (both Katsoulos et al.24 (49%) and Sabesanet al.23 (45%) used materials with considerably lower watercontents); lens thickness (Katsoulos et al.24 (0.12 mm) andChen et al.22 (0.182 mm) used substantially thinner lenses)and optical zone designs (Chen et al.22 designed their cus-tomised zones on the posterior lens surfaces). Compared tothicker designs, thinner lenses may show lower levels of light scattering and other optical effects caused by lens ex-ure, perhaps resulting in improved optical quality. In con-trast to higher water content lenses, lower water contentlenses would show a higher modulus, which may assist intrapping the tear lm beneath the lens, potentially creating

Figure 5. Comparing visual performances measured with the patient’shabitual RGP lenses, the standard toric soft lenses (TCLs) and both the

100% and 50% lenses in the contact lens group, using repeated-mea-sures analysis of variance ( RM-ANOVA). Error bars represent 1 standarddeviation.



8/17/2019 Opo 12133

9/12

a uid lens, which might partially neutralise lower- andhigher-order aberrations. Compared to anterior surfacedesigns, lenses with customised optical zones on the poster-ior lens surface are more likely to match the anterior cor-neal surface topography, potentially minimising on-eyelens rotation and translation. 22

Our ndings regarding reductions in aberrations withthe customised lenses on-eye are in agreement with previ-ous studies. 22,23,25

– 28 Specically, our results show that the100% lenses tended to overcorrect vertical coma, resultingin positive residual aberration. In contrast, the 50%lenses tended to provide either negative residual aberration(contact lens group) or values close to zero (spectaclesgroup). As each ACCL was veried using the ClearWaveaberrometer, our recorded on-eye aberration measure-ments may have been inuenced by small lens decentra-tions, alongside potential spot-imaging errors at thewavefront sensor. 48

– 51

The rationale for using a ‘partial’ correction was basedon previous studies 50 – 56 conrming that decentration of a‘full’ wavefront-guided correction (through either rotationor translation) induces superuous residual aberrations,thereby diminishing visual performance. These reports alsoconrm that decentration of a ‘partial’ correction still yields a greater visual benet compared to conventionalsphero-cylindrical corrections. 52,53,57 Our results, in part,agree with these studies, showing generally lower magni-tudes of residual higher-order aberrations with the 50%lenses than with the 100% lenses; however, these differenceswere not statistically signicant in either study group.

As with the TCLs, the poor visual performances with

either ACCL type, in the contact lens group, may also beattributed to large magnitudes of residual astigmatism, asboth ACCLs induced substantially large magnitudes of 2nd-order cylinder RMS error. The inducement of suchcylindrical error is perhaps attributable to small magnitudesof on-eye lens translation. 55 Similar results were also evi-dent in the spectacle wearers, but to a lesser degree, perhapsdue to dissimilarities in sample size. Such residual cylindri-cal errors appear to impact on visual performance, as sig-nicant correlations were found between 2nd-ordercylinder RMS and both HCVA and LCVA measured witheither ACCL type (Table S2, Supporting Information).

In contrast to RGP lenses, all three study lenses induceda negative shift in spherical aberration in the contact lenswearers. A similar trend was also found in the spectaclewearers, but to a lesser extent. We believe this negative shiftis due to an inherent correction for positive spherical aber-ration, a typical design feature of our study lenses, which isalso found in other customised soft lens designs used fortting irregular corneas. 22,23 The magnitude of negativespherical aberration induced is typically dependent on eachlens’ spherical power and asphericity.

Signicant correlations between residual 3rd-order comaRMS aberrations and LCVA scores, recorded with the TCLsand 50% lenses, indicate that residual coma is more detri-mental to LCVA than HCVA, corroborating previous nd-ings in normal 58 and keratoconic subjects. 37 As the residualcoma aberrations measured with the TCLs on-eye were

used to design the aberration control of both ACCLs, any errors in measuring these residual aberrations may have ledto inaccuracies in determining the ‘ideal’ magnitude of coma correction required, ultimately inuencing the nalvisual performance. 37,59 In support of this theory, the comaRMS aberrations measured with the TCLs in the contactlens group showed a large degree of variability (i.e. largeerror bars in Figure 3).

Furthermore, the substantial residual defocus and cylindri-cal errors measured with all three study lenses were not veri-ed subjectively. In the vast majority of cases, over-refractionrevealed only minor alterations in spherical power to maxi-mise visual performance. These ndings agree with previousreports, 5,24 and are most likely to be due to errors at thewavefront sensor whilst evaluating highly-aberratedeyes.48,49,60 Such errors may have resulted in an underestima-tion or overestimation of the residual aberrations measuredwith the TCLs on-eye.5,24 Consequently, alternative measure-ment techniques, such as high-dynamic range Hartmann-Shack wavefront sensors 61 or laser-ray tracing methods, 62

might provide less variable data in keratoconic patients.When customised ACCLs are applied on-eye, superu-

ous wavefront aberrations can become induced throughsmall lens translations and/or rotations upon blinking. 50,55

However, such lens movements are essential to allow ade-

quate tear exchange. As in other studies,17,22,23,28,50

ourresults showed that small magnitudes of on-eye rotationand translation, between blinks, were unavoidable with ourstudy lenses in a few patients, due to their irregular cornealproles. On-eye translation and/or rotation of a ACCL caninduce superuous 2nd-order and 3rd-order aberrations,respectively, thereby impacting on visual performance. 37

For all participants, the maximum lens rotation (in primary gaze) with the TCLs was found to be ≤ 5 degrees (in only four out of 22 patients). The manufacturer appropriately adjusted the nal cylinder axis of the TCLs and bothACCLs to compensate for such small rotations. The on-eyelens’ movement, centration and coverage measurementsrecorded with the TCLs and both ACCLs indicated that thelens ttings were each clinically acceptable. Although cor-rections for rotations were applied, a signicant limitationof this study is that our ACCLs did not account for any habitual on-eye lens translation (neither vertically nor hori-zontally).

Depending on their magnitude, as well as the eye’s inher-ent wavefront error, superuous aberrations inducedthrough unwanted ACCL decentrations may either enhance



8/17/2019 Opo 12133

10/12

or reduce the effectiveness of the wavefront correc-tion. 50,54,55,63 Such induced, residual aberrations are pro-portional to the amount of displacement as well themagnitude of the displaced aberration. 42,54,55 Vertical andhorizontal translations, respectively, typically induce largerresidual aberrations, compared to rotational displace-

ments.42,50,54,55

In particular, vertical translation (as typi-cally seen with blinks) of a ‘pure’ correction for 3rd-ordervertical coma induces substantial residual 2nd-order spher-ical and cylindrical errors. 54,55,63 Figure 2 demonstrates theinducement of 2nd-order cylinder RMS error in bothgroups with the study lenses, which adversely impacts onoptical quality. 52,55

Furthermore, translation of a lens containing negativespherical aberration induces superuous negative comaaberrations. 42,54 Due to their design, all three of our study lenses showed an inherent level of negative spherical aber-ration; therefore, residual coma RMS aberrations measuredwith either ACCL type on-eye may be related to small lenstranslations.

Another factor inuencing visual performance with theACCLs is the diameter of the customised optical zone andthe patient’s pupil diameter. 50,51 The ACCLs were eachdesigned for a 4-mm pupil diameter; however, the averagepupil size measured whilst recording HCVA was4.1 0.4 mm [ 1 S.D.]. This slight discrepancy high-lights a signicant limitation of customised lenses. 64 Toovercome, future studies should aim to produce custo-mised lenses with bespoke optical zones, which match thepupil sizes of individual patients, particularly as smallerpupil sizes decrease the tolerances of any on-eye lens move-

ment errors.63

Finally, the RGP lens wearers were instructed to leavetheir habitual lenses out ‘overnight’ prior to attendingtheir lens collection appointments, this time period wasconsiderably shorter than the 7-day washout periodobserved prior to the initial tting appointment. This limi-tation was imposed by the ethical approval granted for thisstudy, in that it was deemed unethical to ask these success-ful RGP lens-wearing patients to repeatedly leave theirlenses out.

In conclusion, the two customised ACCLs substantially reduced 3rd-order RMS and higher-order RMS error in bothgroups. In the contact lens group, the RGP lenses providedbetter visual performances than either ACCL. In the specta-cle-wearing subjects, there were no signicant differences invisual performance between the patient’s habitual spectaclesand the three study lenses. The most probable explanationsfor the poor visual performances achieved with the ACCLsare small, on-eye lens translations (where the customisedoptical zone becomes misaligned with the pupil). Futurestudies developing ACCLs for keratoconic patients shouldensure that corrections for both lens rotation and translation

are applied (perhaps to the order of microns) 23 to optimisethe correction of higher-order aberrations.

Acknowledgements

Amit Jinabhai was supported by a PhD studentship from

The College of Optometrists (UK) and UMIP (The Univer-sity of Manchester, UK). We would like to thank all of ourclinical and clerical colleagues at Manchester Royal EyeHospital and The University of Manchester Optometry Clinics for their assistance with patient recruitment.

Disclosure

The authors have no conicts of interests to declare.

References

1. Kennedy RH, Bourne WM & Dyer JA. A 48-year clinical

and epidemiologic study of Keratoconus. Am J Ophthalmol 1986; 101: 267 – 73.

2. Zadnik K, Fink B, Nichols JJ, Yu J & Schechtman K.Between-eye asymmetry in keratoconus. Cornea 2002; 21:671 – 79.

3. Barr JT & Yackels T. Corneal scarring in keratoconus-mea-surements and inuence on visual acuity. Int Contact LensClin 1991; 22: 173 – 5.

4. Kosaki R, Maeda N, Bessho K et al. Magnitude and orienta-tion of Zernike terms in patients with keratoconus. Invest Ophthalmol Vis Sci 2007; 48: 3062 – 68.

5. Jinabhai A, O’Donnell C & Radhakrishnan H. A comparisonbetween subjective refraction and aberrometry-derived

refraction in keratoconus patients and control subjects. Curr Eye Res 2010; 35: 703 – 14.

6. Abdalla YF, Elsahn AF, Hammersmith KM & Cohen EJ.SynergEyes lenses for keratoconus. Cornea 2010; 29: 5 – 8.

7. Downie LE. Predictive value of corneal topography forClearKone hybrid contact lenses. Optom Vis Sci 2013; 90:e191 – 7.

8. Tomalla M & Cagnolati W. Modern treatment options forthe therapy of keratoconus. Cont Lens Anterior Eye 2007; 30:61 – 6.

9. Ye P, Sun A & Weissman BA. Role of mini-scleral gas-per-meable lenses in the treatment of corneal disorders. EyeContact Lens 2007; 33: 111 – 3.

10. Weed KH, MacEwen CJ, Giles T, Low J & McGhee CNJ. TheDundee University Scottish Keratoconus study: demograph-ics, corneal signs, associated diseases, and eye rubbing. Eye2008; 22: 534 – 41.

11. Wagner H, Barr JT & Zadnik K. Collaborative LongitudinalEvaluation of Keratoconus (CLEK) Study: methods andndings to date. Cont Lens Anterior Eye 2007; 30: 223 – 32.

12. Lim N & Vogt U. Characteristics and functional outcomesof 130 patients with keratoconus attending a specialist con-tact lens clinic. Eye 2002; 16: 54 – 9.



8/17/2019 Opo 12133

11/12

8/17/2019 Opo 12133

12/12

46. Sorbara L, Chong T & Fonn D. Visual acuity, lens exure,and residual astigmatism of keratoconic eyes as a functionof back optic zone radius of rigid lenses. Cont Lens Anterior Eye 2000; 23: 48 – 52.

47. Zadnik K & Mutti DO. Contact lens tting relation andvisual acuity in keratoconus. Am J Optom Physiol Opt 1987;64: 698 – 702.

48. Jinabhai A, Radhakrishnan H & O’Donnell C. Repeatability of ocular aberration measurements in patients with keratoc-onus. Ophthalmic Physiol Opt 2011; 31: 588 – 94.

49. Thibos LN. Principles of Hartmann-Shack aberrometry. J Refract Surg 2000; 16: s563 – 65.

50. Lopez-Gil N, Castejon-Mochon JF & Fern andez-Sanchez V.Limitations of the ocular wavefront correction with contactlenses. Vis Res 2009; 49: 1729 – 37.

51. Thibos LN, Cheng X & Bradley A. Design principles andlimitations of wave-front guided contact lenses. Eye Contact Lens 2003; 29: S167 – 70.

52. de Brabander J, Chateau N, Marin G, L opez-Gil N & BenitoA. Simulated optical performance of custom wavefront softcontact lenses for keratoconus. Optom Vis Sci 2003; 80: 637 – 43.

53. Guirao A, Cox I & Williams DR. Method for optimizing thecorrection of the eye’s higher-order aberrations in the pres-ence of decentrations. J Opt Soc Am A Opt Image Sci Vis2002; 19: 126 – 8.

54. Guirao A, Williams DR & Cox I. Effect of rotation andtranslation on the expected benet of an ideal method tocorrect the eye’s higher-order aberrations. J Opt Soc Am AOpt Image Sci Vis 2001; 18: 1003 – 15.

55. Jinabhai A, Charman WN, O’Donnell C & RadhakrishnanH. Optical quality for keratoconic eyes with conventionalRGP lens and simulated, customised contact lens correc-tions: a comparison. Ophthalmic Physiol Opt 2012; 32:200 – 12.

56. Marsack JD, Rozema JJ, Koppen C, Tassignon MJ & Apple-gate RA. Template-based correction of high-order aberra-tion in keratoconus. Optom Vis Sci 2013; 90: 324 – 34.

57. Shi Y, Queener HM, Marsack JD et al. Optimizing wave-front-guided corrections for highly aberrated eyes in thepresence of registration uncertainty. J Vis 2013; 13: 1 – 15:article 8.

58. Fernandez-Sanchez V, Ponce ME, Lara F et al. Effect of 3rd-order aberrations on human vision. J Cataract Refract Surg 2008; 34: 1339 – 44.

59. Okamoto C, Okamoto F, Oshika T, Miyata K & SamejimaT. Higher-order wavefront aberration and letter-contrastsensitivity in keratoconus. Eye 2008; 22: 1488 – 92.

60. Mihashi T, Hirohara Y, Bessho K et al. Intensity analysis of Hartmann-Shack images in cataractous, keratoconic, andnormal eyes to investigate light scattering. Jpn J Ophthalmol 2006; 50: 323 – 33.

61. Pantanelli S, MacRae S, Jeong TM & Yoon G. Characterizingthe wave aberration in eyes with keratoconus or penetratingkeratoplasty using a high-dynamic range wavefront sensor.Ophthalmology 2007; 114: 2013 – 21.

62. Moreno-Barriuso E & Navarro R. Laser Ray Tracing versusHartmann-Shack sensor for measuring optical aberrationsin the human eye. J Opt Soc Am A Opt Image Sci Vis 2000;17: 974 – 85.

63. Bara S, Mancebo T & Moreno-Barriuso E. Positioning toler-ances for phase plates compensating aberrations of thehuman eye. Appl Opt 2000; 39: 3413 – 20.

64. Charman WN & Chateau N. The prospects for super-acuity:limits to visual performance after correction of monochro-matic ocular aberration. Ophthalmic Physiol Opt 2003; 23:479 – 93.

Supporting Information

Additional Supporting Information may be found in theonline version of this article:

Figure S1. Visual performance: Spectacle wearers.Table S1. Corneal data.Table S2. Correlations between visual performance and

aberrations.



opo 12133

Documents