ABSTRACT

Central corneal thickness alterations may cause residualrefractive errors following laser in situ keratomileusis (LASIK).This study reports associations between central corneal thick-ness alterations and residual refractive error following uncom-plicated LASIK. Ninety-one myopic patients with a meanrefractive correction of –3.91 ± 3.2 DS / –0.66 ± 0.3 DCwere evaluated. Central corneal thickness was measuredprior to, during and following surgery and 2 months laterusing ultrasound pachometry. Results indicate increased tissueremoval (94 ± 33 µm; mean ± SD) compared to the nominalNidek value (52 ± 24 µm, P < 0.001).Twenty-four hours laterthe tissue removal was 46 ± 27 µm.There was no associationbetween altered central corneal thickness and ablation depth(r = 0.058, P = 0.454). Central corneal thickness change wasinversely proportional to residual refractive error (r = –0.364,P < 0.01). Increased tissue removal may occur due to rapidstromal dehydration. Central corneal thickness changesbetween 24 h, and 2 months after surgery were constantover a range of ablation depths, which may partly explain thestability of LASIK procedures over a range of corrections.

Key words: central corneal thickness, LASIK refractivesurgery, residual refractive error.

INTRODUCTION

Laser in situ keratomileusis (LASIK) for the surgical correc-tion of refractive errors has gained widespread acceptance,however, there is limited information on refractive stabilityfollowing the procedure. Refractive stability is influenced byalterations in corneal curvature and corneal thickness, andrefractive error can typically take up to 2 or 3 months to sta-bilize after LASIK.1

This study aimed to determine changes in central cornealthickness (CCT) during and following LASIK procedures, tocompare actual ablation depths with the laser algorithm, andto evaluate associations between early operative changes inCCT and initial ablation depth and refractive error.

METHODS

Subjects

Ninety-one consecutive myopic patients (181 eyes) whohad undergone uncomplicated LASIK surgery using theNidek-EC5000 laser (Nidek, Gamagori, Japan) were invitedto participate. The average patient age was 38 ± 6 years(mean ± SD) and 55 patients were female. The FDA-approved LASIK procedure was performed by two experi-enced surgeons and all surgical procedures did not ventureoutside the normal standardized routine. Measurementswere taken during this routine surgical procedure. Writteninformed consent was obtained from subjects prior to thestudy. A microkeratome was used to create an 8.5–9.5 mmdiameter, 100–160 µm thick flap. All measurements weretaken in the afternoon to eliminate normal diurnal variationsin corneal thickness.

Procedures

Central corneal thickness was measured on all eyes inpatients using a TOMEY SP-2000 ultrasound pachometerTomey Corp., Nagoya, Japan) (ultrasound velocity 1640m/s) prior to surgery. The theoretical ablation depth (laseralgorithm) for the desired refractive error was noted. TheNidek EC-5000 laser algorithm is based on geometricaloptics (Munnerlyn’s formula). This algorithm gives the the-oretical corneal ablation depth required to attain emme-tropia, which ranges from 7.2 to 8.0 µm per 0.50 DS ofrefractive change for a 6.5 mm optic and 5.0–5.5 µm per0.50 DS of refractive change for a 5.5 mm optic. Following

Clinical and Experimental Ophthalmology (2000) 28, 185–187

Lens and Cornea

Change in central corneal thickness following laser in situkeratomileusis for myopiaMark H Feltham MOptom1,2 and Fiona Stapleton PhD2

1Vista Laser Eye Clinic, Canberra, Australian Capital Territory and 2Cooperative Research Centre for Eye Research and Technology,University of New South Wales, Sydney, New South Wales, Australia

■ Correspondence: Mark Feltham, Vista Laser Eye Clinic, Second Floor, Colonial Building, 161 London Circuit, Civic, ACT 2608, Australia.

Email: tsf.mhf@dynamite.com.au

removal of the flap, CCT was measured on the collagen bedbefore and after the excimer laser ablation.

During a normal uncomplicated LASIK procedure,common practice is to perform the ablation between 20 and90 s after the lifting of the flap. Where patients had an iden-tical refractive error, and therefore ablation thickness inboth eyes (n = 7), CCT was measured in one eye 20, 60 and90 s after flap removal and the ablation was performed at the95 s time point. In the contralateral eye, the LASIK ablationwas performed 20 s after the flap was lifted. Changes inrefraction, corneal curvature and thickness were noted 24 hafter surgery.

Residual CCT and refractive error was measured after24 h, 1 week, 1 month and 2 months after surgery. The dif-ference between preoperative CCT (less measured ablationdepth) and the CCT after two months was compared. Therefractive error and the ablation optic zone size used toachieve this result were also noted.

Data analysis

Differences between measured and theoretical ablationdepths were compared using a paired t-test. Associationsbetween measured and theoretical thickness were evaluatedusing a product moment correlation. In the majority of subjects (n = 84), different ablation depths were performedon the right and left eyes, hence in all analyses, data fromeach eye were considered independently.

RESULTS

The theoretical central ablation depth was 52 ± 24 µm(mean ± SD) and the mean measured central ablation depth(CCT after flap lift – CCT after ablation) was 94 ± 34 µm (P < 0.001). However, the ablation depth measured at day 1(CCT prior to surgery – CCT at day 1) was 46 ± 27 µm.Theoretical ablation depth and depth measured at day 1were highly correlated (r = 0.93, P < 0.0001).

In the group of seven subjects with equal refractive errorsin the two eyes, the change in CCT at 24 h was not signifi-cantly different between the eye that was untouched for 95 sbefore ablation and the eye that was ablated after 20 s(P > 0.05). The refractive correction achieved after 24 h wasnot significantly different between the two eyes (P = 0.1). Inaddition, there were no corneal topographical abnormalities

at 24 h in the more dehydrated eye. The edges of the dehy-drated flap were subjectively more noticeable compared tothe control eye after 24 h. The change in CCT from 20 to60 s was 25 ± 5 µm (mean ± SD) and at 90 s, 36 ± 17 µm.

The change in total CCT was inversely proportional tothe residual refractive error (RRE) at the 2 month time point(r = –0.364, P < 0.01, Fig. 1). Myopic RRE was associatedwith thicker than expected corneas at 2 months post-operatively, and the reverse was true for the hypermetropicRRE.

There was no relationship between change in CCT anddepth of tissue removal (r = 0.058, P = 0.454, Fig. 2).

Based on a previous study showing reduced regression forsmaller optic zone sizes,2 the majority of ablations (n = 119)were carried out for a 5.5 mm optic size, with a transitionzone of 6.5 mm. Otherwise, the ablation optic zone size was6 mm and 7.0 mm transition zone (n = 30), and 6.5 mmoptic zone size and 7.5 mm transition zone (n = 32). Therewere significant differences in CCT change with optic zonesize (P < 0.05). The change in CCT was greatest for the6.5 > 6.0 > 5.5 mm optic zones. Thickness change mea-sured 90 s after lifting the flap was greater for the larger flapdiameter (9.5 mm vs 8.5 mm; P < 0.01), however, the RRE at 24 h was not affected by flap diameter (P > 0.05).

DISCUSSION

This study has demonstrated that the actual central cornealthickness, immediately post myopic ablation, is thinner thanthat predicted using laser algorithm. This difference was nolonger apparent 24 h after surgery. Based on the difference between the theoretical and the measured abla-tion depth immediately after surgery, for the study samplesize analysed, with α set to 0.05, and β set to zero, thepower of the analysis is 1.000. For the results presented 24 hafter the procedure, using similar criteria for this samplesize, the estimated power is 0.61. For a power of 0.80, toshow a significant difference (P < 0.05) in thickness 24 hafter ablation, the required sample size would be 285 eyes.

Thickness data measured following lifting of the flapwould suggest that the exposed stroma becomes rapidlydehydrated following flap removal. Within the subsequent24 h, there appears to be rehydration of the stroma fromtears, aqueous or irrigation fluid, and the CCT is pre-dictable, based on the laser algorithm.

186 Feltham and Stapleton

Figure 1. Change in total central corneal thickness (µm) withresidual refractive error (dioptres D), 2 months postoperatively.

Figure 2. Change in total central corneal thickness (CCT; µm)with ablation depth (µm), 2 months postoperatively.

The degree of thinning due to dehydration occurring inthe first 90 s after the flap lift does not appear to affect theRRE. A period prior to ablation of up to 90 s appears toresult in a predictable refractive result; however, the samplesize was small in this study, and so was the subsequent powerto detect differences (0.3). In addition, 90 s of exposureprior to the laser ablation did not appear to result in the formation of central islands or cause topographical abnor-malities. As the excimer laser is more effective on a drysurface,3–5 it could be hypothesized that for the ablation tobe even, regardless of the ablation depth, the stromal dehy-dration should be uniform in order to achieve a predictableablation curve.

Central corneal thickness changed in inverse proportionto the RRE, with myopic RRE associated with centralcorneal thickening. In one study of moderate and highmyopia, similar findings were reported 1 year after LASIK,where a RRE of –0.96 D was associated with a change inCCT of 15 µm.6 Regression of the refractive effect inmyopia has also been associated with central thickening dueto epithelial hyperplasia.7 The mechanism for epithelialhyperplasia is unclear, but may involve stimulation ofstromal keratocytes to synthesize specific cytokines, namelygrowth factors which act on epithelial cells (type II sig-nalling) to cause hyperplasia.8 Hepatocyte growth factorand keratinocyte growth factor are paracrine growth factorsproduced by activated keratocytes, which may promoteepithelial thickening by stimulating epithelial proliferationand inhibiting differentiation.9 Measurement of peripheralcorneal thickness corresponding to the maximum depth ofablation in hypermetropic RRE would confirm whether asimilar effect occurred in hypermetropia. This study was

unable to differentiate between a change in CCT due tostromal or epithelial thickness changes, and either optical orhigh frequency ultrasound pachometry would be required toclarify this.

The magnitude of the change in CCT was not propor-tional to the ablation depth, which differs from PRK,10

where greater regression occurs for higher correctionsattempted. These data support the non-occurrence ofgreater regression for higher refractive errors in LASIK.Within this study population there was refractive error sta-bility over a wide range of refractive corrections attemptedduring LASIK procedures.

ACKNOWLEDGEMENTS

The authors would like to thank surgeons Dr Richard Wolfe,FRACO FRACS and Dr Ron Binetter FRACO FRACS, MrReg Wong for statistical advice and Professor Brien Holden.

REFERENCES

1. Cameron JD. Fundamentals of Cornea and External Disease. Phila-delphia: Mosby, 1997.

2. Feltham MH et al. Clin. Exp. Optom. 2000; (in press).3. Trokel SL et al. Am. J. Ophthalmol. 1983; 96: 710–15.4. Krauss JM et al. Sur. Ophthalmol. 1986; 1: 37–53.5. Marshall J et al. Ophthalmology 1985; 92: 749–58.6. Chayet AS et al. Ophthalmology 1998; 105: 1194–9.7. Reinstein DZ et al. Ophthalmology 1999; 106: 474–82.8. Rocha G et al. Int. Ophthalmol. Clin. 1996; 36: 9–20.9. Wilson SE. J. Refract. Surg. 1997; 13: 171–5.

10. Snibson GR et al. Arch. Ophthalmol. 1995; 113: 994–1000.

Corneal thickness change in LASIK 187