Femtosecond laser versus mechanical keratome flaps in wavefront-guided laser in situ keratomileusis: Prospective contralateral eye study

Femtosecond laser versus mechanicalkeratome flaps in wavefront-guided laserin situ keratomileusis

Prospective contralateral eye study

Daniel S. Durrie, MD, Guy M. Kezirian, MD

Purpose: To compare the outcomes of wavefront-guided laser in situ keratomileusis(LASIK) performed using the IntraLase femtosecond laser with the outcomes usingthe Hansatome mechanical microkeratome.

Setting: Private clinic, Overland Park, Kansas, USA.

Methods: In a prospective contralateral-eye study performed under institutionalreview board supervision, 51 consecutive patients (102 eyes) had bilateralwavefront-guided LASIK for myopia using the Alcon LADARVision laser. One eye ofeach patient was randomized to have the flap created with the IntraLasefemtosecond laser and the other flap using a standard compression headHansatomemicrokeratome. All other treatment parameters were the same.

Results: The IntraLase group had significantly better mean uncorrected visualacuity (UCVA) at all intervals from 1 day to 3 months postoperatively. Themean spheroequivalent at 3 months was more myopic with the Hansatome(�0.34 diopter [D] 6 0.28 [SD]) than with the IntraLase (�0.19 6 0.24 D) (P!.01).The mean residual astigmatism at 3 months was also significantly higher inthe Hansatome group than in the IntraLase group (0.32 6 0.25 D and0.17 6 0.20 D, respectively) (P!.01). The differences in UCVA persisted afterspheroequivalent outcomes were controlled for but equilibrated when the analysiswas modified to control for manifest postoperative astigmatism. Aberrometryshowed significantly higher astigmatism and trefoil in the Hansatome group.Recovery of corneal sensation and epithelial integrity was similar between groups.

Conclusions: The statistically better UCVA and manifest refractive outcomesafter LASIK with the IntraLase femtosecond laser may be the result of differencesin postoperative astigmatism and trefoil. These findings are consistent withprevious findings of better astigmatic outcomes with the IntraLase laser and mayhave clinical significance for wavefront-guided treatments.

J Cataract Refract Surg 2005; 31:120–126 ª 2005 ASCRS and ESCRS

Refractive surgery based on aberrometry—wave-

front-guided laser in situ keratomileusis (LASIK)—

attempts to provide superior visual outcomes by

correcting the overall wavefront error. Applegate and

coauthors1 examined the effect of various aberrations on

visual acuity and found that defocus and astigmatism

affect visual acuity more than higher-order aberrations

(HOAs). They conclude that the first goal of wavefront-

guided corrections should be to reliably achieve patient-

� 2005 ASCRS and ESCRS

Published by Elsevier Inc.

specific target correction of sphere and cylinder without

inducing new HOAs.

Previous studies report that the LASIK flap

may induce HOAs2 and advocate the use of photo-

refractive keratectomy (PRK) for wavefront-guided

treatments.3 Others report that flap-induced aberra-

tions vary widely from eye to eye and that most HOAs

observed after LASIK are induced by the ablation, not

the flap.4

0886-3350/05/$-see front matter

doi:10.1016/j.jcrs.2004.09.046

CORNEAL BIOMECHANICS: FLAP INFLUENCE ON LASIK OUTCOMES

In a prospective randomized trial of 18 eyes of

9 patients, Tran5 found that significantly smaller in-

creases in lower-order aberrations and HOAs were

induced by IntraLase flaps than by flaps created with the

Hansatome mechanical keratome (Bausch & Lomb,

Inc.). In that study, the flaps were created, lifted, and

replaced without excimer ablation.

In a retrospective study of 376 eyes having

conventional LASIK, Kezirian and Stonecipher6 found

significantly better astigmatic outcomes with the Intra-

Lase femtosecond laser (IntraLase Corp.) than with

2 mechanical keratomes, the Hansatome and the

Carriazo-Barraquer (Moria, Inc.). Specifically, they

found higher amounts of postoperative astigmatism in

eyes having treatment for spherical refractive errors and

attributed the differences to the flap. They conclude that

better astigmatic neutrality might make the IntraLase

more appropriate to wavefront-guided treatments than

the Hansatome or Carriazo-Barraquer keratomes.

This prospective study evaluated the influence of

the flap creation method on wavefront-guided treat-

ments by comparing the IntraLase femtosecond laser

and the Hansatome mechanical keratome.

Patients and MethodsThe prospective randomized comparative clinical study

was performed under protocol with investigational reviewboard supervision. Enrollment in the study was offered toconsecutive patients having bilateral custom LASIK for thecorrection of myopia with or without astigmatism whoagreed to follow the examination schedule and met thefollowing criteria: a manifest myopic refractive error up to�7.00 diopters (D) sphere with less than �0.50 Dastigmatism in both eyes or aberrometer dilated refractionup to �7.50 D sphere with less than �1.50 D astigmatism.

Accepted for publication November 1, 2004.

FromDurrie Vision (Durrie), Overland Park, Kansas, and SurgiVisionConsultants, Inc. (Kezirian), Paradise Valley, Arizona, USA.

Presented at the ASCRS Symposium on Cataract, IOL and RefractiveSurgery, San Diego, California, USA, May 2004.

Dr. Durrie is a paid consultant to IntraLase Corporation and AlconLaboratories, Inc. Dr. Kezirian received financial compensation fromIntraLase Corporation for his assistance with data analysis andpreparation of the manuscript.

Reprint requests to Guy M. Kezirian, MD, SurgiVision Consultants,Inc., 2183 Hathaway Avenue, Westlake Village, California 91362-5170, USA. E-mail: [email protected].

J CATARACT REFRACT SUR

Additional criteria included a minimum age of 21 years,a normal ophthalmic examination except for refractive error,normal topography, no history of ocular surgery, stablerefraction within 60.50 D over the past year, no ongoing useof ophthalmic or systemic medications, and no history ofautoimmune disease.

In all patients, the flap was created with the IntraLasefemtosecond laser flap in 1 eye and with the Hansatomemechanical keratome flap in the contralateral eye. Theselection of the flap creation method was done at the time ofenrollment using a predefined randomization schedule.

Surgical TechniqueBilateral same-session LASIK procedures were per-

formed in all patients by 1 surgeon (D.S.D.) using 1 excimerlaser (LADARVision 4000, software version Jupiter 5.2) withthe eye tracker. Results from 3 aberrometry measurementswere averaged using the aberrometer software and enteredinto the laser without surgeon adjustments. The targetedoutcome was plano in all cases.

All flaps were created using a superior hinge. Hansatomeflaps were created using a standard compression 180 mmhead, Accuglide blades (Bausch & Lomb), and a 9.5 mmsuction head. IntraLase flap settings were 9.0 mm diameter,118 mm depth, 55-degree hinge, and 65-degree side-cutangle.

After the IntraLase flap was created in 1 eye, the patientsat for 15 minutes before the flap was lifted for ablation.Hansatome flaps were created after a drop of balanced saltsolution (BSS) was placed on the central cornea; the flap wasimmediately lifted and the ablation performed. Pachymetrywas performed before the flap was created, after the flap waslifted, and before the ablation.

ExaminationsPatients were evaluated preoperatively and 1 day, 1 week,

and 1 and 3 months after surgery. Results from the preop-erative and 1- and 3-month postoperative visits are reportedhere. Clinical examinations were performed by optometriststrained in LASIK evaluations who were blinded to which eyehad the femtosecond laser flap.

Soft contact lenses were removed at least 3 days and hardcontact lenses for 3 weeks before the preoperative examina-tion. The preoperative evaluation included measurement ofuncorrected visual acuity (UCVA) using the Early TreatmentDiabetic Retinopathy Study chart (Lighthouse Vision),manifest refraction, and corneal sensitivity as well as slitlampand dilated fundus examinations. A Schirmer test wasperformed after 1 drop of proparacaine 1% ophthalmicsolution was administered. All wavefront measurements weredone with the pupils in a dilated, noncycloplegic stateobtained using 1 drop each of phenylephrine 2.5% andtropicamide 1% solution. Aberrometry was done in all eyespreoperatively and at 3 months.

121G—VOL 31, JANUARY 2005


Outcome MeasuresMajor outcome measures included manifest refraction,

UCVA, best spectacle-corrected visual acuity (BSCVA), andaberrometry data for cylinder, coma, spherical aberration, andtrefoil. Recovery of epithelial integrity and corneal sensationwas also compared.

Subset AnalysisEyes were divided into subsets based on 3-month re-

fractive data in an attempt to determine whether the differ-ences in UCVA could be explained by the differences in thepostoperative refractions. The 4 subsets were all eyes, eyeswith 0.50 D or less postoperative spheroequivalent error, eyeswith 0.25 D or less spheroequivalent error, and eyes with0.25 D or less spheroequivalent error and no manifest refrac-tive astigmatism. In each subset, the postoperative UCVA wascompared with the preoperative BSCVA to avoid influence ofthe preoperative BSCVA on the results.

Statistical AnalysisAnalysis of clinical outcomes was performed using

Refractive Surgery Consultant Elite software (RefractiveConsulting Group, Inc.) and the Analysis Tool Pack of

122 J CATARACT REFRACT SURG

Figure 1. Comparison of UCVA 20/16 (bottom 2 lines) and 20/20

(top 2 lines) rates in the Hansatome group (squares) and IntraLase

group (triangles). Note the slight delay in improvement in the 20/16

rates compared with the 20/20 rates in both groups. Although the rate

of improvement was similar between the 2 groups, the mean acuities

were significantly better in the IntraLase group at each interval

(Student t test).

Table 2. Postoperative refractive results.

1 Week 1 Month

Result Hansatome IntraLase P Value* Hansatome IntraLase P Value*

Spheroequivalent

Mean 6 SD (D) �0.28 6 0.34 �0.12 6 0.27 !.01/NS 0.26 6 0.30 �0.14 6 0.24 !.02 6 !.05

60.50 (%) 75 94 !.01 80 94 !.05

Astigmatism

Mean 6 SD (D) 0.27 6 0.32 0.13 6 0.19 !.01 6 !.001 0.26 6 0.27 0.14 6 0.21 .01 6 !.05

60.25 (%) 65 88 .01 67 88 !.03

60.50 (%) 80 98 !.01 88 96 NS

Mean correction

efficacy† (%) 6 SD

142 6 67 117 6 42 !.03 6 !.001 141 6 90 118 6 35 NS 6 !.001

NSZ not significant

*t test for means; F test for variance (SD); chi-square analysis for rates†Ratio of attempted to achieved cylinder vector

Table 1. Visual outcomes.

Group N Mean P Value*SD

(Lines)20/20 orBetter (%) P Value*

20/16 orBetter (%) P Value*

UCVARBSCVA† (%) P Value*

Hansatome 51 20/17.9!.01

1.1 86!.03

53!.05

71!.05

IntraLase 51 20/16.1 0.8 98 73 86

BSCVAZ best spectacle-corrected visual acuity; NZ number of eyes; UCVAZ uncorrected visual acuity

*t test for means; F test for variance (SD); chi-square analysis for rates†Postoperative UCVA same as or better than preoperative BSCVA

—VOL 31, JANUARY 2005


Microsoft Excel. Statistical tests were performed at the 95%confidence interval unless otherwise noted. The Student t testwas used to compare mean outcomes, the F test to comparevariance, and chi-square analysis to compare rates.

Refractive data were calculated at the corneal vertex. Allvisual acuity calculations were performed using logMARequivalents.7 Vector analysis was used to analyze cylinderoutcomes,8 and doubled-angle plots were calculated usingmethods described by Holladay and coauthors.9 Aberrometryinformation was interpreted using normalized coefficientsaccording to methods described by Thibos10 and Atchison.11

Results

Enrollment, Demographics, and AccountabilityOne hundred two eyes of 51 patients were treated

over the 7-week period of July 9 through August 29,

2003. Because of the odd number of patients (51), there

were 25 right eyes and 26 left eyes in the IntraLase co-

hort and 26 right eyes and 25 left eyes in the Hansatome

group.

Patient demographics for the 2 cohorts were iden-

tical as each patient contributed 1 eye to each cohort.

There were 35 women (69%) and 16 men (31%) with

a mean age of 34.7 years 6 7.7 (SD). The preoperative

refractive data were similar between the groups. The

mean preoperative spheroequivalent and cylinder were

the same in both groups, �3.59 D and �0.64 D,

respectfully.

All patients (51/51, 100%) appeared for the 1- and

3-month visits.

Uncorrected Visual AcuityFigure 1 shows the postoperative distance UCVA

results. The mean acuities in both groups improved

Table 2 (cont).

3 Months

Hansatome IntraLase P Value*

�0.34 6 0.28 �0.19 6 0.24 !.01 6 NS

78 90 NS

0.32 6 0.25 0.17 6 0.20 !.01 6 NS

63 82 .03

88 98 .05

142 6 114 130 6 49 NS 6 .001

J CATARACT REFRACT SURG

significantly from 1 week to 3 months, and the rates of

change were similar in the 2 groups. The mean acuities

were significantly better in the IntraLase group than in

the Hansatome group at each interval (Student t test).

Table 1 shows the 3-month UCVA. The mean

UCVA and standard deviations were significantly better

in the IntraLase group than in the Hansatome group. In

addition, significantly more eyes in the IntraLase group

had a UCVA of 20/16 and 20/20 (P!.03 and P!.05,

respectively; chi-square analysis).

Manifest Refractive SpheroequivalentMore eyes in the IntraLase group than in the

Hansatome group were within60.50 D of the intended

manifest refractive spheroequivalent at 1 week and 1

month (Table 2). At 3 months, the difference was not

statistically significant (90% and 78%, respectively)

(PZ .10).

The slope of the linear regression of programmed

versus achieved spheroequivalent based on the 3-month

data was 0.91 in the Hansatome group (r2 Z 0.96) and

0.95 in the IntraLase group (r2 Z .97). The 4% differ-

ence in slope was reflected in a small but significantly

more myopic mean spheroequivalent in the Hansatome

group at all intervals (Table 2).

AstigmatismThe mean astigmatism results were significantly

better in the IntraLase group than in the Hansatome

group at all follow-up visits (P!.01; Table 2). More

eyes in the IntraLase group than in the Hansatome

group were within 60.25 D of the intended target at all

intervals and within 60.50 D at 1 week and 3 months,

but not at 1 month. The rate of eyes in the IntraLase

group within 60.50 D approached 100% throughout

the follow-up.

The standard deviations of postoperative astigma-

tism and correction efficacy were significantly better in

the IntraLase group than in the Hansatome group at all

intervals. Both groups tended to be overcorrected for

astigmatism. This is shown in Table 2 as the correction

efficacy, which is the ratio of the intended to the

achieved astigmatic vector. At 3 months, the Hansa-

tome group had a mean overcorrection of 42% and the

IntraLase group, of 30%. However, the mean pre-

operative astigmatism was low, so slight overcorrections

may be disproportionately represented.

123—VOL 31, JANUARY 2005


There was no consistent orientation of the post-

operative cylinder axis in either group. The mean astig-

matism increased slightly between 1 week and 3 months

in both groups; however, the differences were not statis-

tically significant.

Subset AnalysisTable 3 shows the results of the analysis of the

4 subsets. As eyes with smaller amounts of residual

postoperative refractive errors were excluded from anal-

ysis, the ratio of eyes achieving a postoperative UCVA

as good as or better than the preoperative BSCVA was

more similar between the IntraLase and Hansatome

groups. Although the smaller cohorts made it more dif-

ficult to show significant differences between groups,

rates favored the IntraLase until eyes with any post-

operative cylinder were excluded, at which point 100%

of Hansatome eyes achieved a postoperative UCVA as

good as or better than the preoperative BSCVA.

Aberrometry MeasurementsThere were no significant differences between the

IntraLase and Hansatome groups in any preoperative

aberration. At 3 months, the results were similar

between the 2 groups except for Z 22 (with-/against-

the-rule astigmatism) and Z �33 (oblique trefoil), both of

which were significantly higher in the Hansatome group

than in the IntraLase group. The manifest refractive

Table 3. Efficacy of UCVA by subset.

Subset Eyes UCVA R BSCVA* P Value

All eyes .05

Hansatome 51 71%

IntraLase 51 86%

60.50 D .07

Hansatome 40 78%

IntraLase 46 91%

60.25 D .08

Hansatome 26 81%

IntraLase 37 95%

60.25 D/no cyl NS

Hansatome 11 100%

IntraLase 25 96%

BSCVA Z best spectacle-corrected visual acuity; no cyl Z no

cylinder; NSZ not significant; UCVAZ uncorrected visual acuity

*Postoperative UCVA same as or better than preoperative BSCVA


astigmatism did not correlate with the aberrometry

astigmatism measurements in either group (Figure 2).

The aberrometry findings of higher postoperative

astigmatism are consistent with the refractive find-

ings. The postoperative mean aberrometry root mean

square (RMS) for Z 22 was 0.152 6 0.232 mm in the

Hansatome group and 0.028 6 0.233 mm in the

IntraLase group (P!.01).

Figure 3 shows the postoperative RMS distribution

for trefoil. The Hansatome eyes had a significantly

higher postoperative mean and standard deviation RMS

values for trefoil than the IntraLase eyes. There was

a slight correlation between postoperative RMS trefoil

and postoperative astigmatism in the Hansatome group

(0.24) but not in the IntraLase group (0.02).

Other MeasuresThere was no significant difference between the

2 groups in the rate of recovery of corneal sensation and

epithelial integrity.

DiscussionThis prospective contralateral-eye study compared

LASIK outcomes obtained using 2 devices for flap cre-

ation: the IntraLase femtosecond laser and the standard

compression head Hansatome mechanical keratome.

Overall, results in both groups were excellent, with 88%

of Hansatome eyes and 98% of IntraLase eyes achieving

a UCVA of 20/20 or better at 3 months. The UCVA

differences were most apparent at the 20/16 level (54%

Hansatome versus 73% IntraLase) and in the percentage

of eyes achieving a postoperative UCVA as good as or

better than the preoperative BSCVA (71% and 86%,

respectively).

All eyes were treated based on aberrometry measure-

ments without surgeon adjustments. Surgeon adjust-

ment of the sphere might have improved the outcomes

in both groups. The slope of the linear regression of the

programmed versus the achieved spheroequivalent was

0.91 in the Hansatome group and 0.95 in the IntraLase

group.

Postoperatively, there was slightly greater resid-

ual spheroequivalent in the Hansatome group (mean

�0.34 6 0.28 D) than in the IntraLase group (mean

�0.19 6 0.24 D); however, this only partially explains

the differences in UCVA. When eyes with residual



Figure 2. Scatterplot of postoperative manifest refractive astigmatism versus cylinder measured with aberrometry in the Hansatome group

(left) and the IntraLase group (right). The correlation was poor in both groups, suggesting that what is measured clinically as manifest cylinder may

be displayed as other aberrations (eg, trefoil) by the aberrometer.

spheroequivalent errors were removed from the analysis,

the differences in UCVA became less pronounced.

However, the differences equilibrated only when eyes

with residual astigmatism were removed from analysis.

This suggests that residual astigmatism may have played

a significant role in the differences in UCVA between

the 2 groups.

Results confirm a previous report6 of better mani-

fest astigmatism outcomes with the IntraLase laser. The

Hansatome group had higher amounts of postoperative

RMS astigmatism, which agrees with the clinical find-

ings of higher amounts of manifest astigmatism in these

eyes. However, cylinder measurements obtained at the

aberrometer did not correlate well with clinical mea-

surements of manifest astigmatism in either group. This

Figure 3. Histogramdistribution of the postoperative RMS trefoil in

the Hansatome group (light bar) and IntraLase group (dark bar). Both

the mean and standard deviation were significantly better in the

IntraLase group. The preoperative values were similar.

J CATARACT REFRACT SUR

may suggest that the higher amounts of trefoil in the

Hansatome group contributed to the differences in

manifest astigmatism.

Although nomogram adjustments may have im-

proved the spheroequivalent results in both groups, the

increases in astigmatism and trefoil in the Hansatome

group would probably not be improved with nomogram

adjustments. This factor may make the IntraLase laser

a better option than the Hansatome microkeratome in

the treatment and prevention of HOAs after LASIK.

Future studies of flap biomechanics may help

explain the differences in our refractive and aberrometer

findings. Additional comparative studies of PRK and

femtosecond laser–assisted LASIK could help determine

which treatment modality is best for wavefront-guided

treatments.

References1. Applegate RA, Sarver EJ, Khemsara V. Are all aberrations

equal? J Refract Surg 2002; 18:S556–S5622. Pallikaris IG, Kymionis GD, Panagopoulou SI, et al. In-

duced optical aberrations following formation of a laserin situ keratomileusis flap. J Cataract Refract Surg 2002;28:1737–1741

3. Panagopoulou SI, Pallikaris IG. Wavefront customizedablations with the WASCA Asclepion workstation. J Re-fract Surg 2001; 17:S608–S612

4. Porter J, MacRae S, Yoon G, et al. Separate effects ofthe microkeratome incision and laser ablation on theeye’s wave aberration. Am J Ophthalmol 2003; 136:327–337

5. Tran D. Influence of flap on higher order aberrationsafter LASIK. In press, J Refract Surg

125G—VOL 31, JANUARY 2005


6. Kezirian GM, Stonecipher KG. Comparison of the Intra-Lase femtosecond laser and mechanical keratomes forlaser in situ keratomileusis. J Cataract Refract Surg 2004;30:804–811

7. Holladay JT. Proper method for calculating average vi-sual acuity. J Refract Surg 1997; 13:388–391

8. Holladay JT, Cravy TV, Koch DD. Calculating the sur-gically induced refractive change following ocular sur-gery. J Cataract Refract Surg 1992; 18:429–443


9. Holladay JT, Moran JM, Kezirian GM. Analysis of ag-gregate surgically induced refractive change, predictionerror, and intraocular astigmatism. J Cataract RefractSurg 2001; 27:61–79

10. Thibos LN. Wavefront data reporting and terminology.J Refract Surg 2001; 17:S578–S583

11. Atchison DA. Recent advances in representation ofmonochromatic aberrations of human eyes. Clin ExpOptom 2004; 87:138–148


Documents

Femtosecond laser versus mechanical keratome flaps in wavefront-guided laser in situ keratomileusis: Prospective contralateral eye study