relationship between corneal topographical changes and subjective myopic reduction in overnight...

cxo_489 237..242

RESEARCH PAPER

Relationship between corneal topographical changes andsubjective myopic reduction in overnight orthokeratology:

a retrospective study

Clin Exp Optom 2010; 93: 4: 237–242 DOI:10.1111/j.1444-0938.2010.00489.x

Ben Chan BSc (Hons) OptomPauline Cho PhD FAAO FBCLAJohn Mountford Dip App Sc FCLSAFAAOSchool of Optometry, The Hong KongPolytechnic University, Hong Kong SAR,ChinaE-mail: [email protected]

Background: To investigate the relationship between the change in the manifest refrac-tive error (DM), the change in apical corneal power (DACP) and initial corneal asphe-ricity (Q) in overnight orthokeratology (ortho-K).Methods: One hundred and twenty-eight clinical records of children undergoingortho-K from a university optometry clinic were reviewed. The refractive and topographi-cal data at baseline and at two-week visit of 58 patients who fulfilled the inclusion criteriawere retrieved and analysed.Results: Significant differences (p < 0.001) between the change in manifest refractiveerror and changes in the apical corneal power or the maximum change in corneal power(DMCP) within the treatment zone were found. Linear regression analysis was used todescribe the change in manifest refractive error and the change in apical corneal power,and the change in manifest refractive error and the maximum change in corneal power,with the equations: DM = 0.91DACP + 0.57 (r = 0.78, p < 0.001) and DM = 0.93DMCP +0.01 (r = 0.79, p < 0.001) respectively. On average, the change in apical corneal powerunderestimated the change in manifest refractive error by 0.34 � 0.57 D; whereas onaverage, the maximum change in corneal power overestimated the change in manifestrefractive error by 0.23 � 0.57 D (paired-t-tests, p < 0.001). A low but significant corre-lation between initial corneal asphericity and the change in manifest refractive error(Spearman r = -0.33, p = 0.01) was observed.Conclusions: The change in apical corneal power underestimates the change in manifestrefractive error in ortho-K, whereas the maximum change in corneal power overestimatesthis parameter. Compared with retinoscopy and autorefraction, the change in apicalcorneal power is still useful for estimation of the change in manifest refractive error.Although the maximum change in corneal power appears to give a closer estimation ofthe change in manifest refractive error than the change in apical corneal power, there isno advantage in the use of maximum corneal power (manually located) instead of apicalcorneal power (a default given by the topographer) to estimate the change in manifestrefractive error, as there is no significant difference in the estimations by either param-eter. Initial corneal asphericity measured by the Medmont E300 corneal topographer haslimited usage in predicting the change in manifest refractive error in overnight ortho-K.

Submitted: 22 November 2009Revised: 11 March 2010Accepted for publication: 18 March 2010

Key words: apical corneal power, corneal asphericity, corneal topography, myopic reduction, orthokeratology

C L I N I C A L A N D E X P E R I M E N T A L

OPTOMETRY

© 2010 The Authors Clinical and Experimental Optometry 93.4 July 2010

Journal compilation © 2010 Optometrists Association Australia 237

Corneal topography is an essential tech-nique in orthokeratology (ortho-K) prac-tice. It is used to provide lens fitting data aswell as in the objective assessment of over-night fitting. Corneal topographical mapsalso provide useful information aboutcorneal surface abnormalities, such askeratoconus, corneal astigmatism andapex decentration, which are contra-indications for ortho-K.1 Axial and tangen-tial topographical maps allow assessmentof the ortho-K lens performance on thecornea during sleep. The difference mapfeature is specifically useful for monitoringlens centration and patient managementduring the procedure. Munnerlyn’s for-mula2 was originally used to calculate theablation depth in photorefractive keratec-tomy to determine the refractive change.The formula has also been used to accountfor refractive changes in ortho-K. Mount-ford3 suggested that myopic reduction inortho-K can also be determined from thechange in the apical corneal power (ACP)calculated from the axial difference map.He conducted a retrospective study exam-ining both refractive changes and changesin corneal parameters generated by theEyeSys 2000 (version 3.2 aspheric algo-rithms) corneal topographer in 60 subjectsafter one month of ortho-K lens wear.Analysis revealed a high correlationbetween changes in the apical cornealpower and refractive changes (refractivechange = 0.92DACP + 0.15, r = 0.95) and hesuggested using the change in apicalcorneal power as an objective method forestimating myopic reduction following theprocedure.

Because ortho-K involves corneal shapechange, various initial corneal parametershave been suggested as predictors for theoutcome of the procedure.3–7 A factor thatcan accurately predict the outcome of thetreatment would help practitioners tobetter estimate the likely level of success ina particular patient before commencingthe procedure. Wlodyga and Bryla6 sug-gested using the difference betweencentral and temporal keratometric read-ings as a means for predicting the degreeof refractive change after ortho-K,however, this parameter has not provenuseful.5,8,9 Lui and Edwards7 found that the

initial central corneal thickness had themost predictive value, that is, thickercorneas achieved greater myopic reduc-tion in daywear ortho-K. Other parametersincluding pretreatment corneal shapeindex (shape factor [p], eccentricity value[e] or asphericity [Q]) as a possible pre-dictive factor have been suggested.3,5,10

Some of these studies drew this conclusionfrom the result of the positive correlationbetween change in corneal shape indexand myopic reduction rather than usingthe initial shape index.3,11,12 In fact, the useof post-ortho-K corneal shape index pro-vided by the corneal topographer is inap-propriate, as the cornea no longerresembles a conic section. Although theinitial e and p values have been reportedto be significantly correlated with overallmyopic reduction in ortho-K by Joe andcolleagues5 and Lui and Edwards,10 bothstudies used a day wear protocol andrecruited small numbers of subjects (n =11 and 28, respectively).

While a normal cornea is oftendescribed in terms of a conic section,according to the Baker’s equation, apicalcorneal radius and eccentricity (e2) areneeded to characterise the conic sectionof the cornea; however, e2 may have anegative value, when the conic section hasan oblate shape, in which case e has nomeaning. To interpret all possible shapesthat can occur in a normal cornea, insteadof using e, an alternative corneal index,such as asphericity, to describe the shapeof the cornea, which is capable of describ-ing both prolate and oblate cornealshapes (Q = -e2) has been suggested.13 Thisparameter can be accurately and repro-ducibly provided by the Medmont E-300corneal topographer (Medmont Pty Ltd,Camberwell, Australia).14–16 Nichols andassociates12 reported that a change incorneal asphericity was correlated with asignificant myopic reduction after 60 daysof overnight ortho-K treatment, however,this study observed the altered topographyafter ortho-K rather than using initialcorneal asphericity as a predictor forortho-K.

This study aimed to investigate the rela-tionship between changes in centralcorneal power and manifest refractive

changes. We also aimed to determine ifthe initial meridional corneal asphericityis a good predictor of the outcome ofortho-K in terms of percentage of myopicreduction.

METHODS

A total of 128 consecutive ortho-K patientfiles in the Optometry Clinic of The HongKong Polytechnic University were retro-spectively analysed. The non-cycloplegicsubjective refraction recorded at the visitsbefore commencement of ortho-K treat-ment (baseline) and after two-weeks oflens wear was retrieved. As there were nostatistically significant changes in subjec-tive refractive cylinder during the two-week lens wear period (mean change 0.04� 0.54 D; paired-t-test, p > 0.05), only sub-jective refractive sphere was used for analy-sis. Achieved myopic reduction (DM) wasthe difference between the subjectiverefractive spheres at baseline and at thetwo-week visits. The inclusion criteria arelisted in Table 1. Only data for the righteye were analysed. Of the files reviewed, 58patients fulfilled the inclusion criteria.Table 2 presents the demographic data ofthese patients. All patients used eithereLens (E&E Optics Asia Ltd, Hong Kong)or DriemLens design (Taiwan MacroVision Group, Taiwan). The two lensdesigns were similar (four-curve reversegeometry) and were fitted according tothe guideline from the manufacturer withthe back optic zone radius fitted on flat K+ 0.75.

The relationship between the achievedmyopic reduction and the change inapical corneal power was evaluated. Theapical corneal power was derived from thecentre of the topographic map usingthe refractive index of 1.3375. Because themaximum change in corneal power maynot necessarily be the default apicalcorneal power given by the topographer,we also investigated the relationshipbetween the achieved myopic reductionand maximum change in corneal powerwithin the treatment zone (DMCP). Thetreatment zone was determined by movingthe cursor from the centre of the corneato the points on the nasal and temporal

Corneal changes and myopic reduction in orthokeratology Chan, Cho and Mountford

Clinical and Experimental Optometry 93.4 July 2010 © 2010 The Authors

238 Journal compilation © 2010 Optometrists Association Australia

sides, where the difference between thepre- and post-treatment corneal refractivepowers was zero. Both the change in apicalcorneal power and the maximum changein corneal power were determined fromthe difference topographical map (axialsubtractive plot) between baseline and thetwo-week visit.

To investigate whether initial cornealasphericity could be used for predictingthe ortho-K effect, the flattest cornealasphericity at 9.80 mm chord was used. Asthe same amount of myopia can be associ-ated with a different corneal asphericity,percentage myopic reduction (%DM) wasused to investigate the correlation be-

tween initial corneal asphericity and theachieved myopic reduction.

To detect a minimum difference of 0.25D between the achieved myopic reductionand the change in apical corneal power, asample size of 41 subjects is required toproduce a power of 80 per cent (a = 0.05,2-tailed).

Treatment of dataStatistical analyses were performed usingSPSS 15.0 (SPSS Inc, Chicago, USA). Datanormality was tested using Kolmogorov-Smirnov D-tests before statistical analysis.A p-value of 0.05 was set to indicate statis-tical significance. All data, except the per-centage myopic reduction were normallydistributed; therefore, paired-t-tests wereused to test for differences betweenthe achieved myopic reduction and thechange in apical corneal power and themaximum change in corneal power;Spearman correlation coefficient was usedto determine the relationship betweeninitial corneal asphericity and the percent-age myopic reduction.

RESULTS

The pre- and post- ortho-K corneal andrefractive cylinder information is shown inTable 3. A significant difference was foundbetween the achieved myopic reductionand the change in apical corneal powerafter the two weeks of lens wear, with theachieved myopic reduction greater thanthe change in apical corneal power by amean of 0.34 � 0.57 D (paired-t-test, p <0.001) and there was a significant correla-tion between the two variables (Pearson r= 0.78, p < 0.001) (Figure 1). Their rela-tionship can be described by the equationof DM = 0.91DACP + 0.57.

A significant mean difference of 0.61 �

0.47 D was found between the change inapical corneal power and the maximumchange in corneal power (paired-t-test, p <0.001). A significant difference was foundbetween the means for the achievedmyopic reduction and the maximumchange in corneal power after two weeksof lens wear; however, the mean achievedmyopic reduction was smaller than the

Age between 10 and 15 years.First time orthokeratology lens wearer.At least six hours of lens wear on the night before the two-week visit.The two-week visit was between 9 am and 12 noon.Available clinical data included non-cycloplegic refraction and bull’s eye subtractive

topographical pattern measured by the Medmont E300 corneal topographer (version 4.8.0,Medmont Pty Ltd, Camberwell, Australia).

Achieved at least 50 per cent of the attempted myopic change on the day of examination.The same pair of lenses was worn for at least six months at the time of data retrieval.

Table 1. Subject inclusion criteria

Female/male n = 39/19

Median RangeMean � SD

Age (years)* 10 10 to 15Baseline refractive sphere (D) -3.38 � 1.38Baseline refractive cylinder (D)* -0.50 0 to -2.25Baseline corneal astigmatism (D) -1.19 � 0.57Average myopic reduction (D) 2.91 � 0.91

Table 2. Demographical data of the patients (n = 58) (*non-Gaussiandistribution)

Baseline Post-wear Change

ACP (flat meridian) (D) 43.58 � 1.19 41.02 � 1.12 2.57 � 0.79MCP (flat meridian) (D) 43.90 � 1.19 40.37 � 1.17 3.18 � 0.87Refractive sphere (D) -3.38 � 1.38 -0.47 � 0.89 2.91 � 0.91Refractive astigmatism (D) -0.53 � 0.54 -0.50 � 0.58 0.04 � 0.40Corneal astigmatism (D) -1.19 � 0.57 -1.23 � 0.53 -0.03 � 0.53

ACP: apical corneal powerMCP: maximum corneal power

Table 3. The pre- and post-orthokeratology corneal and refractive powers (n = 58)




mean maximum change in corneal powerof 0.23 � 0.57 D (paired-t-test, p < 0.001).The relationship between the achievedmyopic reduction and the maximumchange in corneal power was DM =0.93DMCP + 0.01 (Pearson r = 0.79, p <0.001) (Figure 2).

At the two-week visit, a significant butweak correlation was found between initial

corneal asphericity and the percentagemyopic reduction (Spearman r = -0.33, p =0.01) (Figure 3).

DISCUSSION

The achieved myopic reduction was sig-nificantly associated with the change inapical corneal power but the latter signifi-

cantly underestimated the former by anaverage 0.34 � 0.57 D. The relationshipbetween the two parameters can be repre-sented by the equation of DM = 0.91DACP+ 0.57 (r = 0.78) (Figure 1). ThoughMountford3 had previously reported ahigher correlation (r = 0.95) between thechange in apical corneal power and theachieved myopic reduction with a relation-ship of DM = 0.92DACP + 0.15, the differ-ence in the results between this study andour study may be due to the methodolo-gies and corneal topographers used.Mountford3 used data from the visit, whichshowed the maximum refractive changewithin a six-month study period, whereasin the present study, data were collectedafter two weeks lens wear, irrespective ofthe refractive errors.

The method used for the reconstruc-tion of the apical corneal power also dif-fered as the EyeSys corneal topographerwas used in Mountford’s study. This instru-ment uses a basic spherical reconstructionalgorithm and applies an aspheric bestfit to the data to calculate the value ofthe apical corneal power,17 whereas theMedmont corneal topographer used inthe current study has an arc-step recon-struction algorithm and then extrapolatesinto the centre to determine the apicalcorneal power.18 The difference in corneal

ACP (D)

M (D

)

0 1 2 3 4 5

5

4

3

2

1

0

y = 0.91x + 0.57r = 0 .78

Figure 1. Relationship between achieved myopic reductionand the change in apical corneal power (DACP) (n = 58)

MCP (D)

M (D

)

0 1 2 3 4 5

5

4

3

2

1

0

y = 0.93x + 0.01r = 0 .79

Figure 2. Relationship between achieved myopic reductionand maximum change in corneal power (DMCP) (n = 58)

140

120

100

80

60

40

20

0-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0

y = -34.06x + 58.46r = - 0.33

%M

(D)

Initial Q

Figure 3. Relationship between percentage of change inmyopic refraction and initial Q (asphericity)




curvature reconstruction algorithmsbetween the two instruments may result ina different apical corneal power measure-ment after ortho-K.16

As shown in Figure 1, the averagemyopic reduction was 0.34 D greater thanthat indicated by the change in apicalcorneal power. This may be due to practi-tioners tending to under-correct ortho-Kpatients when determining residual refrac-tion to emphasise the possible effect ofthe treatment. The difference betweenthe achieved myopic reduction and thechange in apical corneal power may alsobe due to the method of calculation of theapical corneal radius by the Medmontcorneal topographer after ortho-K. Theapical radius is calculated by extrapolatingthe flat and steep meridians within adefault machine-defined area. Anotherexplanation of the discrepancy may beerrors in the calculation of the apicalradius following an abrupt change of thecornea after ortho-K.

In the current study, we did not analysechanges in refractive astigmatism as therewere no significant changes in astigmatismafter ortho-K lens wear. Because this is aretrospective study, we did not have inclu-sion criteria for the refraction (spherical orcylindrical powers) but only two subjectshad refractive astigmatism over 2.00 D andboth showed no significant post-ortho-Kchange in their astigmatism. A comprehen-sive analysis of changes in refractive astig-matism and corneal toricity in ortho-K inthis group of subjects using the Thibosvector analysis is described elsewhere.19

The amount of change in apical cornealpower (measured with the MedmontE-300 corneal topographer) could notaccount for the total myopic reduction fol-lowing ortho-K probably because theremay be other factors affecting changes incorneal shape (D asphericity), which affectthe optics of the eye and visual perception.A previous study has also shown that objec-tive methods using retinoscopy andautorefraction underestimate the changein refractive error after ortho-K by 0.20 �

0.42 D and 0.54 � 0.62 D, respectively.20 Inaddition, the change in apical cornealpower is a measure of the power change atthe apex of the cornea, which is a point.

On the other hand, subjective refractionmeasures the refractive error of the totalocular system, which involves a greaterarea of the central corneal surface, whichis dependent on the size of the post-ortho-K treatment zone, as well as thepupil size. The apical corneal power mea-sured from the centre of the topographicmap is usually offset from the pupil centreand this may also have a potential influ-ence on the discrepancy between thechange in apical corneal power and theachieved myopic reduction. The use of arefractive index of 1.3375 in determiningthe apical corneal power could alsoaccount for some of the differencesbetween corneal and refractive changes.For example, a change in corneal curva-ture from 7.8 mm to 8.2 mm would lead toa change in corneal power of 2.11 D (thatis, from 43.27 D to 41.16 D), if a refractiveindex of 1.3375 were used (that is, thekeratometric refractive index used by mostcorneal topographers for axial power) or achange of 2.35 D (that is, from 48.21 D to45.85 D), if the ‘true’ corneal refractiveindex of 1.376 were used.

In the current study, the maximumchange in corneal power gave a closer esti-mate (mean difference between themaximum change in corneal power andthe achieved myopic reduction of 0.22 D)of the amount of myopic reduction com-pared to the use of the change in apicalcorneal power (mean difference betweenthe change in apical corneal power andthe achieved myopic reduction of -0.34 D);however, the difference did not reachclinical significance. Also, because thechange in apical corneal power is given bythe topographer by default, whereas MCPhas to be located manually by the user,there is no advantage in using MCPinstead of ACP to estimate the achievedmyopic reduction after overnight ortho-K.

Although our results showed that thechange in apical corneal power is not aclose estimation of the achieved myopicreduction after ortho-K lens wear, thechange in apical corneal power is still rec-ommended as an objective method toreflect change in myopic reduction, as it issimilar to the findings by retinoscopy andbetter than autorefraction.20 Also, the

change in apical corneal power is availablefrom subtractive topographic plot.

Although many studies have suggestedinitial ocular parameters such as subjec-tive refraction,21 keratometry read-ings,6,10,22 corneal thickness7 andintraocular pressure5,23 as predictive toolsfor myopic reduction in ortho-K, none ofthe these parameters has been confirmedto be useful for predicting ortho-K out-come.5,21,24 Some studies3,5,10 also reporteda significant correlation between cornealshape index (Q) and myopic reductionfollowing ortho-K, however, the post-ortho-K corneal shape is no longer anormal conic sectional surface. Thecorneal shape descriptive indexes (p,e-value or Q) given by the corneal topog-rapher are invalid for describing suchcorneal shape. Therefore, we examinedthe usefulness of initial corneal aspheric-ity for the prediction of percentagemyopic reduction in ortho-K. The resultshowed a significant but weak relation-ship between initial corneal asphericityand the percentage of myopic reduction(r = -0.33, p = 0.01). Therefore, initialcorneal asphericity is not useful topredict the percentage of myopic reduc-tion in ortho-K. In addition to the knowl-edge, skill and experience of the ortho-Kpractitioner, a reliable parameter, whichcan accurately predict the amount ofreduction of myopia after ortho-K is stillneeded. Further research is required todefine such a factor.

CONCLUSIONS

Change in apical corneal power underes-timates the achieved myopic reduction by0.34 D in ortho-K. In comparison to retin-oscopy and autorefraction, it is still auseful method to provide an estimate ofthe expected post-ortho-K refractivechange. Although the maximum changein corneal power appears to give a closerestimation of the achieved myopic reduc-tion than the change in apical cornealpower, there is no significant differencein the estimations by either parameter.Hence, there is no advantage to usemaximum change in corneal powerinstead of the apical corneal power to esti-




mate the achieved myopic reduction.Initial corneal asphericity is not useful as apredictive factor for the outcome inortho-K.

ACKNOWLEDGEMENT

The results of this study were presented atthe 2nd Asia Cornea & Contact Lens Con-ference, April 2005, and the 15th Asia-Pacific Optometric Congress, October2005.

GRANTS AND FINANCIAL SUPPORT

This study was supported by a grant fromthe School of Optometry, The Hong KongPolytechnic University (A356).

REFERENCES

1. Mountford J, Ruston D, Dave T, Cornealtopography and its measurement. In: DaveT, ed. Orthokeratology: Principles andPractice, chap 2. London: Butterworth-Heinemann, 2004.

2. Munnerlyn CR, Koons SJ, Marshall J. Pho-torefractive keratectomy: a technique forlaser refractive surgery. J Cataract RefractSurg 1988; 14: 46–52.

3. Mountford J. An analysis of the changes incorneal shape and refractive error inducedby accelerated orthokeratology. ICLC 1997;24: 128–144.

4. Coon LJ. Orthokeratology. Part II: Evaluat-ing the Tabb method. J Am Optom Assoc1984; 55: 409–418.

5. Joe JJ, Marsden HJ, Edrington TB. The rela-tionship between corneal eccentricity andimprovement in visual acuity with orthok-eratology. J Am Optom Assoc 1996; 67: 87–97.

6. Wlodyga RJ, Bryla C. Corneal molding: theeasy way. Contact Lens Spectrum 1989; 4:58–65.

7. Lui WO, Edwards MH. Orthokeratology inlow myopia. Part 1: Efficacy and predictabil-ity. Contact Lens Anterior Eye 2000; 23: 77–89.

8. Erickson P, Thorn F. Does refractive errorchange twice as fast as corneal power inorthokeratology? Am J Optom Physiol Opt1977; 54: 581–587.

9. Soni PS, Horner DG. Orthokeratology. In:Bennett ES, Weissman BA, eds. ClinicalContact Lens Practice. chap 49. Philadel-phia: JB Lippincott, 1993.

10. Lui WO, Edwards MH. Orthokeratology inlow myopia. Part 2: Corneal topographicchanges and safety over 100 days. ContactLens Anterior Eye 2000; 23: 90–99.

11. El Hage SG, Leach NE, Shahin R. Con-trolled kerato-reformation (CKR): an alter-native to refractive surgery. Practical Optom1999; 10: 230–235.

12. Nichols JJ, Marsich MM, Nguyen M, BarrJT, Bullimore MA. Overnight orthokeratol-ogy. Optom Vis Sci 2000; 77: 252–259.

13. Swarbrick H. Mind your Ps and Qs! 2ndGlobal Orthokeratology Symposium. 2004.Toronto, Canada.

14. Cho P, Lam AK, Mountford J, Ng L. Theperformance of four different cornealtopographers on normal human corneasand its impact on orthokeratology lensfitting. Optom Vis Sci 2002; 79: 175–183.

15. Chui WS, Cho P. A comparative study of theperformance of different corneal topogra-phers on children with respect to orthok-eratology practice. Optom Vis Sci 2005; 82:420–427.

16. Tang W, Collins MJ, Carney L, Davis B. Theaccuracy and precision performance offour videokeratoscopes in measuring testsurfaces. Optom Vis Sci 2000; 77: 483–491.

17. Roberts C. Characterization of the inherenterror in a spherically-biased corneal topog-raphy system in mapping a radially asphericsurface. J Refract Corneal Surg 1994; 10: 103-111.

18. Riley C. Who’s Who in Corneal Topogra-phy. Contact Lens Spectrum 2004; 5: 58–65.

19. Cheung SW, Cho P. Astigmatic changes inorthokeratology. Optom Vis Sci 2009; 86:1352–1358.

20. Cheung SW, Cho P. Subjective and objec-tive assessments of the effect of orthokera-tology; a cross-sectional study. Curr Eye Res2004; 28: 121–127.

21. Carkeet NL, Mountford JA, Carney LG. Pre-dicting success with orthokeratology lenswear: a retrospective analysis of ocularcharacteristics. Optom Vis Sci 1995; 72: 892–898.

22. Freeman RA. Predicting stable changes inorthokeratology. Contact Lens Forum 1978; 3:21–31.

23. Kerns RL. Research in orthokeratology.Part VIII: Results, conclusions and discus-sion of techniques. J Am Optom Assoc 1978;49: 308–314.

24. Lu F, Jiang J, Qu J, Jin W, Mao X, Shen Y.Clinical study of orthokeratology in youngmyopic adolescents. ICLC 1999; 26: 113–116.

Corresponding author:Dr Pauline ChoSchool of OptometryThe Hong Kong Polytechnic UniversityHong Kong SARCHINAE-mail: [email protected]




relationship between corneal topographical changes and subjective myopic reduction in overnight...

Documents