Femtosecond Sub-Bowman Keratomileusis: A Prospective, Long-term, Intereye Comparison of Safety and...

s

ws6nrPcrtmf(mr

a

(.U(a(.ad

5

Femtosecond Sub-Bowman Keratomileusis: AProspective, Long-term, Intereye Comparison of Safety

and Outcomes of 90- Versus 100-�m Flaps

GAURAV PRAKASH, AMAR AGARWAL, DHIVYA ASHOK KUMAR, MATHANGI CHARI, ATHIYA AGARWAL,
SOOSAN JACOB, AND DHRUV SRIVASTAVA
0atthm.

sso©

I

hwst

● PURPOSE: To analyze the long-term safety profile,visual and refractive results, and incidence of complica-tions between sub-Bowman keratomileusis with 90- and100-�m flaps.● DESIGN: Prospective, randomized, comparative clinicaltudy.

● METHOD: A total of 385 candidates (770 eyes) under-ent bilateral, single-sitting, sub-Bowman keratomileu-

is, with flap creation (90 or 100 �m) on IntraLase0-kHz (Abott Medical Optics) and ablation on Tech-olas 217z100 (Technolas PV) . Right and left eyes wereandomized to undergo 90- or 100-�m flap procedures.reoperative and postoperative assessment included un-orrected visual acuity (UCVA), best spectacle-cor-ected visual acuity (BSCVA), refraction, andopographic analysis. All cases were followed up until 12onths after surgery. After excluding cases lost to

ollow-up, a final analysis of 368 patients was carried out368 eyes in each of the 2 groups). The main outcomeeasures were BSCVA, UCVA, complication rates, and

esidual spherical equivalent refractive error.● RESULTS: The mean preoperative values were: spheri-cal equivalent, �6.08 � 2.7 diopters (D; 90-�m group)nd �5.99 � 2.8 D (100-�m group; P � .7); and

logarithm of the minimal angle of resolution BSCVA,0.01 � 0.03 (90-�m group) and 0.01 � 0.04 (100-�mgroup: P � .8). Postoperative 12-month values were:spherical equivalent, �0.02 � 0.4 D (90-�m group) and�0.01 � 0.4 D (100-�m group; P � .8); logarithm ofthe minimal angle of resolution BSCVA, �0.05 � 0.0790-�m group) and �0.04 � 0.07 (100-�m group; P �8); and logarithm of the minimal angle of resolutionCVA, 0.012 � 0.01 (90-�m group) and 0.017 � 0.02

100-�m group; P � .2). No loss of BSCVA was seen inny case. The efficacy indices were 1.039 � 0.2190-�m group) and 1.014 � 0.24 (100-�m group; P �2); safety indices were 1.163 � 0.21 (90-�m group)nd 1.158 � 0.22 (100-�m group; P � .6); and visionifference indices were 0.09 � 0.14 (90-�m group) and

Accepted for publication Mar 10, 2011.From Dr Agarwal’s Eye Hospital and Eye Research Centre, Chennai,

India (G.P., Am.A., D.A.K., M.C., At.A., S.J., D.S.).Inquiries to Amar Agarwal, Dr Agarwal’s Eye Hospital and Eye

Research Centre, 19, Cathedral Road, Chennai 600 086, India; e-mail:
[email protected]
© 2011 BY ELSEVIER INC. A82

.10 � 0.15 (100-�m group; P � .1). Both groups hadlow but comparable incidence of diffuse lamellar kera-

itis and microstriae. However, the incidence of micros-riae (although visually asymptomatic) was significantlyigher in ablation with spherical equivalent of �9 D orore compared with lesser ablations (6.7% vs 0.8%; P <

001).● CONCLUSIONS: The 1-year follow-up of femtosecondub-Bowman keratomileusis with 90- and 100-�m flapsuggests that both the flap options have comparableutcomes. (Am J Ophthalmol 2011;152:582–590.

2011 by Elsevier Inc. All rights reserved.)

N RECENT YEARS, THERE HAS BEEN A SIGNIFICANT IM-

provement in our understanding of the alterations ofthe cornea after excimer laser ablation. It has been

suggested that surface ablation procedures have betterbiomechanical outcomes compared with thick, flap-basedprocedures.1 However, traditional surface ablation is asso-ciated with slower recovery and increased risk of haze,limiting its useful in higher amounts of ablation.2–4 A

ybrid approach is sub-Bowman keratomileusis. The termas proposed by Slade and associates in a comparative

tudy between thin-flap (100 �m) laser in situ kera-omileusis (LASIK) and photorefractive keratectomy.5

Sub-Bowman keratomileusis combines the faster visualrecovery of flap-based LASIK with the biomechanicalbenefits of a surface ablation. A sub-Bowman flap currentlycan be made by either a microkeratome or a femtosecondlaser, both having good results in either short- or interme-diate term follow-up.6–9

An important factor in determining the stability of athin flap is the amount of epithelium and stroma in it.Reinstien and associates showed that the corneal epithelialthickness is approximately 53.4 � 4.6 �m at the vertex.10

Therefore, a sub-Bowman flap should have a sufficientamount of anterior stromal collagen to prevent tenting,leading to microstriae formation; this is more so afterdeeper ablations.11,12

It has been shown in previous studies that a femtosecond90-�m or 100-�m flap can be created with comparablearchitectural quality and similar early visual results com-pared with thicker flaps on the same platform.9,13 Some
surgeons prefer not to make the 90-�m flap because of a
LL RIGHTS RESERVED. 0002-9394/$36.00doi:10.1016/j.ajo.2011.03.019

mailto:[email protected]

ve

fCa(kr

spds

mtddflApe

potential risk of a flap that is too thin, interface haze, orboth. However, to the best of our knowledge, there are nocomparative long-term studies (1 year or longer) of sub-Bowman keratomileusis options (90 and 100 �m) in aprospective, intereye comparative fashion. Therefore, inthis study, we analyzed the long-term safety profile,visual and refractive results, and incidence of flapcomplications (diffuse lamellar keratitis [DLK], micros-triae, or haze) and the long-term complications (ectasiaor regression) in a fellow-eye comparison between 90-and 100-�m flaps.

METHODS

THIS PROSPECTIVE, RANDOMIZED, COMPARATIVE CLINICAL

study was conducted at a tertiary care and teachingophthalmic hospital. The initial cohort consisted of young,myopic patients who sought treatment at our refractivesurgery service for bilateral femtosecond-assisted LASIK.All patients underwent a detailed LASIK assessment,including anterior segment slit-lamp microscopy, cornealtopographic analysis by Orbscan IIz (Bausch & Lomb,Rochester, New York, USA), wavefront assessment byZywave aberrometry (Bausch & Lomb), biometry, anddilated retinal indirect ophthalmoscopy. Exclusion criteriawere: abnormal corneal topography suggestive of formefruste keratoconus or corneal ectasia, minimum cornealthickness less than 480 �m (Orbscan measurement withacoustic factor of 0.94), residual bed thickness less than300 �m at a 6.5-mm optical zone, local or systemiccontraindications for LASIK, candidates who declined toundergo sub-Bowman keratomileusis (and instead optingfor a 120-�m flap), and candidates who declined to befollowed up for at least 12 months. A total of 385candidates were included and underwent sub-Bowmankeratomileusis. A computer-generated random numberbinary option sequence was used to decide the flap thick-ness. In this, each patient was allotted a random value bythe software, where if the value was 1, the right eye waskept as 90 �m and the left eye was kept as 100 �m, and ifalue was 0, the right eye was kept as 100 �m and the leftye was kept as 90 �m.

All patients underwent flap creation by a 60-kHzemtosecond laser (IntraLase; Abott Medical Optics,hicago, Illinois, USA). To prevent intersurgeon vari-

tion, all cases were performed by the same surgeonG.P.) and same assistant (D.S.). The flap diameter wasept between 8.5 and 8.7 mm in all patients. Theemaining parameters were: flap thickness, 100 or 90 �m;

superior hinge; hinge angle, 55 degrees; bed energy, 0.9�J; spot separation, 7 �m; line separation, 7 �m;ide-cut energy, 1.5 �J; side-cut angle, 70 degrees;ocket enable, on; pocket width, 0.250 �m; pocket startepth, 220 �m; and both pocket tangent and radian spot
eparation, 5 �m.
COMPARATIVE OUTCOMES OF FEMTOSECOVOL. 152, NO. 4

Stromal ablation was performed with the Technolas217z100 laser (Technolas PV, St Louis, Missouri, USA)with iris recognition-based Zyoptix personalized (wave-front-guided) algorithms. The stromal bed was irrigatedwith balanced salt solution after the excimer laser ablationto remove any debris, and then the flap was repositioned.Complications, if any, were noted. After surgery, patientsreceived ofloxacin 0.3% eyedrops 3 times daily and dexa-methasone 0.1% eyedrops 3 times daily for 1 week.Carboxymethyl cellulose drops were given 8 times daily for1 month and were tapered to as and when requiredafterward.

Postoperative evaluation consisted of uncorrected visualacuity (UCVA); best spectacle-corrected visual acuity(BSCVA); refraction; and slit-lamp assessment at 1 day, 1week, 1 month, 3 months, 6 months, and 12 months aftersurgery. Topography was performed at every postoperativevisit beginning in month 1. Anterior segment opticalcoherence tomography (Visante; Carl Ziess AG, Jena,Germany) was carried out at 1 month after surgery using aflap thickness tool with manual override to analyze flapthickness. The mean flap thickness outcomes at each ofthe zones (from the corneal vertex, � is in positive and –is in the negative x-axis direction of the image): � 0.5

m, � 1.5 mm, � 2.5 mm, � 3.5 mm was calculated alonghe 4 axes (0 degrees ¡ 180 degrees, 45 degrees ¡ 225egrees, 90 degrees ¡ 270 degrees, 135 degrees ¡ 315egrees), as described previously.13 The observer analyzingap thickness was blinded to the attempted flap thickness.

single observer (D.A.K.) measured all flaps. Seventeenatients lost to follow-up between months 1 and 6 werexcluded from our analysis.

● STATISTICAL ANALYSIS: Because this was a bilateraleye study, the term cases denotes individual eyes and theterm patient denotes each patient (2 eyes each). Alloutcome data were entered into a Microsoft Excel spread-sheet (Microsoft, Redmond, Washington, USA), and sta-tistical calculations were performed using SPSS softwareversion 16.0 for Windows (SPSS, Inc, Chicago, Illinois,USA). The 2-tailed Student t test was used to evaluatecomparisons refractive and visual outcomes for parametricdata and Rank Sum test was used for non parametricdistributions. The P value was considered significant at lessthan 0.05. Standard refractive surgery reporting graphswere used to present the refractive outcome data.14

Visual acuity was converted to logarithm of the mini-mum angle of resolution (logMAR) from the decimalnotation for statistical analysis, using a conversion chart.The logMAR and decimal notations have scales in reversedirection to each other , for example 20/20 is 1.0 inDecimal and 0.0 in logMAR . All the previous literaturewe reviewed on the safety and efficacy indices tookDecimal scores and therefore higher the index, better wasthe outcome. This would become reversed if we took the
logMAR score for index calculation (a lower value will be
ND SUB-BOWMAN KERATOMILEUSIS 583

o

better) and thus this would make it difficult for the resultsto be compared. The decimal notations were convertedfrom the respective Early Treatment Diabetic Retinopathyvalues. logMAR is the standard method for comparingstatistical data for visual acuity outcome because it con-verts the geometric sequence of a traditional chart to alinear scale. Thus it was used for statistical comparisonbetween means of the visual acuities in this study too.

The decimal visual outcomes at 12 months were alsoused to derive three indices , first two previously used forreporting refractive surgery outcomes 15 and a third newne:

Efficacy Index � UCVAPostop 12 months ⁄ BSCVAPreop

TABLE 1. Comparison of the Refractive anand at the 1-, 3-, and 6-Month Follow-u

Femtosecond Sub-

90-�m Flaps

Mean Standard

Preoperative

Spherical –5.35 2

Cylinder –1.46 0

SE –6.08 2

CDVA 0.01 0

1 Month after surgery

Spherical 0.10 0

Cylinder –0.14 0

SE 0.06 0

UDVA 0.03 0

CDVA –0.03 0

3 Months after surgery

Spherical 0.02 0

Cylinder –0.11 0

SE –0.01 0

UDVA 0.02 0

CDVA –0.04 0


Spherical 0.01 0

Cylinder –0.10 0

SE –0.02 0

UDVA 0.02 0

CDVA –0.05 0


Spherical 0.01 0

Cylinder –0.09 0

SE –0.02 0

UDVA 0.01 0

CDVA –0.05 0

CDVA � corrected distance visual acuity; SE �

visual acuity.

Sphere, cylinder, SE are in diopters. UDVA

resolution values.aStudent t test.

Safety Index � BSCVAPostop 12 months ⁄ BSCVAPreop

AMERICAN JOURNAL OF584

Visual Difference Index� (BSCVAPostop 12 months – UCVAPostop 12 months)/

BSCVAPostop 12 months

The efficacy index looks at the overall efficacy of theprocedure in terms of change in the working visual acuity,that is, preoperative corrected (BSCVA) and postopera-tive uncorrected (UCVA); the safety index looks at theoverall safety of the procedure evaluating if there is loss ofcorrected visual acuity after the surgery.15

However, the main aim of the LASIK is to reduce thedifference between UCVA and BSCVA. The difference inthe postoperative uncorrected vision and corrected visionafter surgery is an important determinant of the outcome of

sual Outcomes at Preoperative Evaluationr the 2 Groups (90 �m and 100 �m) ofan Keratomileusis

100-�m Flaps

P Valueation Mean Standard Deviation

–5.28 2.6 .7

–1.43 0.9 .6

–5.99 2.8 .7

0.01 0.04 .8

0.06 0.03 .1

–0.14 0.2 .6

0.04 0.0 .1

0.04 0.07 .2

–0.04 0.06 .7

0.01 0.3 .8

–0.10 0.2 .7

0.001 0.4 .9

0.03 0.03 .2

–0.04 0.06 .9

0.01 0.3 .8

–0.10 0.2 .9

–0.01 0.4 .2

0.03 0.2 .2

–0.04 0.06 .7

0.00 0.3 .8

–0.09 0.2 1.0

–0.01 0.4 .8

0.02 0.02 .2

–0.04 0.07 .8

erical equivalent; UDVA � uncorrected distance

DVA are in logarithm of the minimal angle of

d Vips foBowm

Devia

.5

.9

.7

.03

.03

.2

.4

.08

.06

.3

.2

.4

.04

.07

.3

.2

.4

.1

.07

.3

.2

.4

.01

.07

sph

and C

a patient after LASIK. The new index (visual difference

OPHTHALMOLOGY OCTOBER 2011

gm

m(

abf

ge

index) tries to quantify this parameter as a function of theoverall postoperative corrected vision. Compared withefficacy and safety index, where a higher value is better, alower value of visual difference index suggests a betteroutcome.

For example, if a Patient A has a UCVA of 20/40 and aBCVA of 20/20 after LASIK, the aim of the LASIKsurgery is better met than in the case of Patient B, with aUCVA of 20/80 and BSCVA of 20/20. The visual differ-ence index is 0.5 for Patient A and 0.75 for Patient B (thelower result is the better one).

RESULTS

● DEMOGRAPHIC DATA: The mean age for the studyroup was 24 � 3.6 years. There were 205 females and 163ales.

● PREOPERATIVE REFRACTIVE ERROR AND BEST SPEC-

TACLE-CORRECTED VISUAL ACUITY: The mean preop-erative spherical error was �5.35 � 2.5 D (range, �1 to�10 D) for the 90-�m group and �5.28 � 2.6 D (range,�1 to �10.50 D) for 100-�m group (P � .7, t test). The

ean preoperative cylindrical error was �1.46 � 0.9 Drange, �0.25 to -3 D) for the 90-�m group and �1.43 �

0.9 D (range, �0.25 to �3.25 D) for the 100-�m group(P � .6, t test). The mean preoperative spherical equiva-lent was �6.08 � 2.7 D (range, �1.12 to �11.5 D) for90-�m group and �5.99 � 2.8 D (range, �1.12 to �11.25D) for 100-�m group (P � .7, t test). The mean logMAR

FIGURE 1. Scattergram showing attempted versus achievedrefraction at 12 months of the 2 groups of patients undergoingfemtosecond sub-Bowman keratomileusis: those with 90-�mflaps and those with 100-�m flaps. Solid line denotes that theattempted and achieved refraction were the same, and dashedblack lines denote a limit a difference of � 1.0 diopters.

BSCVA was 0.01 � 0.03 (range, 0.0 to 0.2) for the 90-�m �


group and 0.01 � 0.04 (range, 0.0 to 0.2) for the 100-�mgroup (P � 0.8, t test).

● REFRACTIVE AND VISUAL OUTCOMES: The refractivend visual outcomes for the 12-month follow-up are givenelow. Additionally, the results for the 1-, 3-, 6-monthollow-ups are given in Table 1.

Refractive Outcome and Predictability. All eyes were tar-eted for emmetropia. The mean postoperative sphericalquivalent at 12 months was �0.02 � 0.4 D (range, �1.25

to 1.12 D) for the 90-�m group and �0.01 � 0.4 D (range,

FIGURE 2. Bar graph showing postoperative spherical equiv-alent refraction at 12 months of the 2 groups of patientsundergoing femtosecond sub-Bowman keratomileusis: thosewith 90-�m flaps and those with 100-�m flaps. CDVA �corrected distance visual acuity; UDVA � uncorrected dis-tance visual acuity.

FIGURE 3. Bar graph demonstrating change in postoperativebest spectacle-corrected visual acuity (BSCVA) at 12 months,compared with preoperative BSCVA, of the 2 groups ofpatients undergoing femtosecond sub-Bowman keratomileusis:those with 90-�m flaps and those with 100-�m flaps. CDVA �corrected distance visual acuity.

1.25 to 1.25 D) for the 100-�m group (P � 0.8, t test).


a9Tf3ge.

eo3flrlato

TgUfam��

wm0f

The number of eyes within � 0.5 D of spherical equivalentt the 12-month follow-up was 350 (95.1%) of 368 for0-�m flaps and 346 (94.02%) of 368 for 100-�m flaps.he number of eyes within � 1.0 D at the 12-month

ollow-up was 365 (99.1%) of 368 for the 90-�m group and60 (97.8%) of 368 for the 100-�m group (Figure 1). Bothroups had comparable spherical equivalent refractiverror over all the follow-ups over the course of 1 year (P �05, all follow-ups, t test).

Defocus Equivalent Refractive Outcome. The number ofyes with a defocus equivalent refractive outcome of 0.5 Dr less at the 12-month follow-up were 232 (63.04%) of68 for 90-�m flaps and 239 (64.9%) of 368 for 100-�maps. The number of eyes with a defocus equivalentefractive outcome of 1 D or less at the 12-month fol-ow-up improved to 353 (95.9%) of 368 for 90-�m flapsnd 250 (95.1%) of 368 for 100-�m flaps. All the eyes inhe 2 groups had a defocus equivalent refractive outcomef 2 D or less at the 12-month follow-up.

Visual Acuity Outcome. In terms of individual Earlyreatment Diabetic Retinopathy values, all eyes in the 2roups had a BSCVA of 20/30 or better before surgery. ACVA of 20/20 or better was seen in 302 (82.06%) of 368

or 90-�m flaps and 298 (80.9%) of 368 for 100-�m flapst 12 months (Figure 2). At the 12-month follow-up, theean logMAR BSCVA was �0.05 � 0.07 (range, 0.0 to0.8) for 90-�m flaps and �0.04 � 0.07 (range, 0.0 to

FIGURE 4. One-month anterior segment optical coherence tompatients undergoing femtosecond sub-Bowman keratomileusimeasurement was carried out using the caliper tool: (Top row

0.8) for 100-�m flaps (P � .8, t test). No loss of BSCVA r


as seen in any of the 3 groups (Figure 3). The 12-monthean logMAR UCVAs were 0.012 � 0.01 (range, �0.8 to

.6) for 90-�m flaps and 0.017 � 0.02 (range, �0.8 to 0.6)or 100-�m flaps (P � .2, t test).

● VISUAL OUTCOME INDICES: The mean efficacy indexof the procedure was 1.039 � 0.21 for 90-�m flaps and1.014 � 0.24 for 100-�m flaps (P � .2, rank-sum test).The mean safety index of the procedure was 1.163 � 0.21for 90-�m flaps and 1.158 � 0.22 for 100-�m flaps (P � .6,

aphy-based assessment of the flap thickness of the 2 groups ofose with 90-�m flaps and those with 100-�m flaps. Theifferent 90-�m flaps; (Bottom row) 2 different 100-�m flaps.

TABLE 2. Anterior Segment Optical CoherenceTomography-Based Assessment of Flap Thickness at Set

Distances from the Corneal Vertex for the 2 Groups ofFemtosecond Sub-Bowman Keratomileusis: 90-�m and

100-�m Flap Groups

Distance from Center of

Corneal Vertex (mm)a

90-�m Flap Group,

Mean � Standard

Deviation (�m)

100-�m Flap Group,

Mean � Standard

Deviation (�m)

� 0.5 90.1 � 6.7 100.6 � 6.9

� 1.5 90.8 � 6.4 100.3 � 6.6

� 2.5 90.6 � 7.7 100.4 � 6.9

� 3.5 90.7 � 6.7 100.8 � 6.5

aDistance was calculated automatically by the caliper tool.

Distance left of the vertex is denoted as –, and distance right of

the vertex is denoted by �.

ogrs: th) 2 d

ank-sum test). The mean vision difference index was 0.09 �


m1sm(ra

0.14 for 90-�m flaps and 0.10 � 0.15 for 100-�m flaps(P � .1, rank-sum test).

● FLAP THICKNESS OUTCOMES: Figure 4 illustrates theeasurement methodology for 1 flap each from the 90- and

00-�m flap groups using the caliper tool on anterioregment optical coherence tomography. The mean flapeasurements were close to the attempted flap thickness

Table 2); however, individual measurements on the flapsanged from 76 to 102 �m for the 90-�m attempted flapnd 85 to 114 �m for the 100-�m flaps.

● COMPLICATIONS: There were no complications duringflap creation or dissection in either group (no flap tears,vertical buttonholes, or uncut areas). No cases of full-thickness flap folds or flap dislodgment were seen. No casesof clinically significant interface haze after 2 weeks of

FIGURE 5. Slit-lamp clinical photographs of different cases o100-�m flaps. (Top left) Clinical photograph of a case witpostoperative day 1. (Top right) Clinical photograph of a casepostoperative week 1. (Bottom) Clinical photograph of a 1-yea

follow-up or transient light sensitivity were seen in either


group. No cases had clinical or topographic evidence ofpost-LASIK ectasia at 12 months.

Diffuse Lamellar Keratitis. DLK (Figure 5, Top left) wasseen in 9 (2.4%) of 368 eyes in the 90-�m flap group andin 12 (3.2%) of 368 eyes in the 100-�m flap group (P � .6,Fisher exact test). Of these, grade I DLK was seen in 7 eyesin the 90-�m group compared with 9 eyes in the 100-�mgroup; and grade II DLK was seen in 2 eyes (0.5%) in the90-�m group compared with 3 eyes (0.8%) in 100-�mgroup. Four patients had bilateral grade 1 DLK, accountingfor 8 of the 21 eyes. All cases of DLK responded well toincreased frequency of topical steroids. No patients hadgrade III or more severe DLK.

Microstriae. Microstriae evident on broad-beam illumi-nation on the slit lamp on the first postoperative day or

tosecond sub-Bowman keratomileusis with attempted 90- or100-�m flap with microstriae and interface debris seen ona 90-�m flap with diffuse lamellar keratitis and microstriae atstoperative uneventful 90-�m flap.

f femh awithr po

first week of follow-up (Figure 5, Top right) were seen in 6


(0sn1f.Dgtp.

aaooict

s(a9tes

speegBat

or

stcasaTpoVeWba

aswMic

of 368 (1.6%) cases of the 90-�m group and in 9 of 368(2.4%) cases in the 100-�m group (3 patients had bilateralmicrostriae, accounting for 6 of 15 (66.7%) cases). Ten ofthese 15 cases had a spherical equivalent (SEQ) of �9 Dor more. In the study, a total of 149 of 736 (20.2%) caseshad an SEQ of �9 D or more. Therefore, the incidence ofmicrostriae in this subgroup (SEQ � �9 D) was 6.7%10/149), which was significantly higher compared with.8% (5/587) of the complimentary (� �9 D SEQ)ubgroup (P � .001, Fisher exact test). However, there waso difference in the microstriae rate in the 90- and00-�m groups in the high-ablation group (� �9 D; 4/70rom the 90-�m group and 6/79 for the 100-�m group; P �7, Fisher exact test) or in the low-ablation group (� �9; 2/298 for the 90-�m group and 3/289 for the 100-�m

roup; P � .7, Fisher exact test). Age, gender, cornealhickness, and eye (right or left) were not found to beredictive for incidence of microstriae in either group (P �

05 for all comparisons).Although none of these patients were visually symptom-

tic, all of them underwent a flap relift and repositioningccording to the protocol of our hospital for managementf early microstriae. Qualitative reduction in the amountf microstriae was seen in all patients, without any changen the UCVA or BSCVA. All the other cases had nolinical abnormalities in the flap morphologic featureshrough 1 year of follow-up (Figure 5, Bottom).

DISCUSSION

SUB-BOWMAN KERATOMILEUSIS IS EVOLVING QUICKLY AS

an alternative for conventional thick-flap excimer proce-dures. In a comparative analysis, Chang noted very lowcomplication rates with 90- and 100-�m flaps.9 In anotherrecent randomized control trial looking primarily at theflap architecture, we found that the 90-, 100-, 110-, and120-�m flaps could be created safely on a current 60-kHzfemtosecond platform, with comparable predictability andplanarity indices in a 1-month follow-up.14 These 2 studiesuggest that parameters measuring flap creation abilityplanarity, predictability, intraoperative complications)nd initial comparative results of follow-up (1 month) of0- and 100-�m flaps are similar. In continuation of thisrend on sub-Bowman flap options, the current studyxamined the long-term outcomes of the 2 types ofub-Bowman flaps in an intereye comparison.

Throughout the follow-up, both groups maintainedimilar refractive and visual outcomes. Good refractiveredictability, UCVA and BSCVA, and high safety andfficacy indices were seen in both groups. The refractiverror stabilized by the third month of follow-up in bothroups. Both groups had comparable changes in theSCVA after surgery. Spherical and defocus equivalentslso were comparable between the 2 groups. All the eyes in
he 2 groups had a defocus equivalent refractive outcome t

f 2 D or less at the 12-month follow-up, suggesting lowesidual astigmatism.

Our results are comparable with those of other results ofub-Bowman keratomileusis in terms of safety outcomes. Inerms of uncorrected visual acuity outcomes (efficacy), ourandidates had results better than those of Azar andssociates in a retrospective, comparative analysis betweenub-Bowman keratomileusis and thick-flap LASIK.16 Azarnd associates performed LASIK in some cases on aechnolas 217z (some of which were done on a Zyoptixlano algorithm, which does not use iris recognition) andthers on a different excimer platform (VISX S2 and S4;ISX, Inc, Santa Clara, California USA). On another

xcimer platform (LADARVision; Alcon Laboratories, Fortorth, Texas, USA), Slade and associates achieved results

etter than those in the current study in terms of percent-ge of cases reading at 20/20 or better.5 However, these 3

studies (Azar and associates, Slade and associates, and thecurrent study) used the same femtosecond platform (In-traLase 60-kHz) to create the sub-Bowman flaps, andtherefore the variation in results seems to arise from theexcimer platform or ablation algorithms, and not from thecreation of sub-Bowman flaps.

The safety and efficacy indices in the 2 groups werecomparable. The results of the visual difference index weregood, with low values in both groups that were statisticallysimilar to each other. Because the visual difference index isa new one, there is no specific range of good or badoutcome with it. However, because the results of thecurrent study are comparable with those of others in theliterature, it can be suggested logically that the range ofvalues of the visual difference index seen in this study canbe an approximate benchmark level for other future studiesusing the visual difference index.

The 2 major complications seen in our 2 groups wereDLK and microstriae formation. The incidence of DLKwas not correlated with the amount of ablation or with theflap thickness. In our study, we also found that microstriaeformation can happen at a significantly higher incidence incases of sub-Bowman keratomileusis for higher (� �9 D)ablations. Although none of our cases were visually symp-tomatic because of the microstriae, surgeons should becareful while handling thin flaps for higher ablation. Therewas no difference between the 90- and the 100-�m flap inthis regard, either.

The major benefit of sub-Bowman keratomileusis oversurface ablation is the presence of a flap that reduceskeratocyte activation and subsequent postablation fibro-blastic transformation and haze.2,17 However, it can bergued that the ablation of sub-Bowman keratomileusis istill in the same corneal zone as in surface ablation, andith a very thin flap, postablation haze can be a possibility.cCulley and Petroll showed that femtosecond laser can

nduce significant keratocyte activation, which may lead tolinical observations of haze in some patients.18 However,
hey recommended that this activation can be avoided by

sapaGh

hphomflthosa

akai

s7bvnafs

g1suThiiditoaa

using lower-raster energies and an extended steroid treat-ment regimen.18

In a retrospective study, Rocha and associates evaluatedcases with varying degree of interface haze after sub-Bowman keratomileusis. The mean achieved flap thicknessin cases with interface haze was 81.65 � 8.9 �m (central)on anterior segment optical coherence tomography at thethird postoperative month.19 Despite the higher lightcattering in these cases, the final visual acuity was notffected. In another study, Kymionis and associates re-orted results of 10 eyes treated with the Schwind Carri-zo-Pendular microkeratome (Schwind eye-tech-solutionsmbH & Co.KG, Kleinostheim, Germany) 90-�mead.20 Flap thickness measured during surgery was as-

sessed as being less than 70 �m in all eyes. All these casesad a transient mild subepithelial haze at the first month’sostoperative examination. A significant improvement inaze formation and residual refractive error were observedn follow-up. The cases with haze in the 2 above-entioned studies have the common factors of undercut

aps (flap thickness much thinner than planned), a flaphickness of approximately 70 to 80 �m, and a moderate toigher amount of refractive error. There was no incidencef visually significant haze, either, in our group, nor in thetudies by Slade and associates, Chang, or Azar andssociates (all 4 on a femtosecond platform).5,9,16 In a

recent refractive surgery consult discussion, Chang noted avery low incidence of interface haze in (3 of 4000 cases ofattempted 90-�m flap) and hypothesized that use oflower-energy parameters could have been the reason for
the low incidence of interface haze in his cases.21 In c
eye study comparing thin-flap LASIK (sub-Bowman’s kera-


nother study on the Schwind Carriazo-Pendular micro-eratome 90-�m head, the same study with haze quotedbove, Kymionis and associates noted that there was noncidence of corneal haze in 47 eyes of 26 patients.22 The

mean flap thickness in this study was thicker (79.88 � 6.94�m) compared with less than 70 �m in all cases in theabove study with interface haze.20 This also stronglyuggests a possible predictive role of flap thickness less than0 to 80 �m in the incidence of interface haze. There maye other possible factors like age, wound-healing responseariations, depth of ablation, variation of epithelial thick-ess (racial or interindividual), and use of antifibroblasticgents that additionally may alter the incidence of inter-ace haze after sub-Bowman keratomileusis, and a futuretudy should look into these factors.

Accidental ultrathin flap formation can lead to verticalas breakthrough and haze, and an attempted 90- or00-�m flap still can be a strong predictive factor. Femto-econd flap laser should be set at a planned depth metic-lously by the technician during the tuning of the laser.herefore, there is a strong role of strict temperature andumidity control, regular cleaning of optics, and standard-

zed docking procedures that may increase the predictabil-ty of thin flaps. The use of careful dissection techniquesuring the flap lift also may prevent damage to the flap,mproving the visual quality and reducing the complica-ion rate. In conclusion, the current study evaluated 2ptions of flap thickness for sub-Bowman keratomileusis inprospective intereye comparison and found that both 90-nd 100-�m flaps created by femtosecond laser have good,
omparable results.
THE AUTHORS INDICATE NO FINANCIAL SUPPORT. DR AMAR AGARWAL IS A CONSULTANT FOR BAUSCH & LOMB, STAARSurgical, and Abbott Medical Optics. None of the other authors have any commercial or propriety interests to disclose. Involved in Design of study(G.P., Am A.); Conduct of study (G.P., D.S., M.C.), Collection (G.P., D.S., D.A.K.), management (Am.A, At.A., S.J.), and analysis and interpretation(G.P., D.A.K.) of data; Administrative, technical, and logistic support (Am.A., At.A., S.J.); and Preparation (G.P., M.C.) and review (Am.A., D.A.,G.P.) of article. Informed consent was obtained from all patients. The study was approved by the Academics and Research Board, Dr Agarwal’s EyeHospital and Eye Research Centre, Chennai, India.

REFERENCES

1. Kamiya K, Shimizu K, Ohmoto F. Comparison of thechanges in corneal biomechanical properties after photore-fractive keratectomy and laser in situ keratomileusis. Cornea2009;28(7):765–769.

2. Moller-Pedersen T, Cavanagh HD, Petroll WM, Jester JV.Stromal wound healing explains refractive instability andhaze development after photorefractive keratectomy: a1-year confocal microscopic study. Ophthalmology 2000;107(7):1235–1245.

3. Carr JD, Patel R, Hersh PS. Management of late corneal hazefollowing photorefractive keratectomy. J Refract Surg 1995;11(3 Suppl):S309–S313.

4. Trattler WB, Barnes SD. Current trends in advanced surfaceablation. Curr Opin Ophthalmol 2008;19(4):330–334.

5. Slade SG, Durrie DS, Binder PS. A prospective, contralateral

tomileusis) with photorefractive keratectomy. Ophthalmol-ogy 2009;116(6):1075–1082.

6. Li H, Sun T, Wang M, Zhao J. Safety and effectiveness ofthin-flap LASIK using a femtosecond laser and microkera-tome in the correction of high myopia in Chinese patients. JRefract Surg 2010;26(2):99–106.

7. Chen HJ, Xia YJ, Zhong YY, Song XL, Chen YG. Anteriorsegment optical coherence tomography measurement offlap thickness after myopic LASIK using the Moria oneuse-plus microkeratome. J Refract Surg 2010;26(6):403–410.

8. Aslanides IM, Tsiklis NS, Astyrakakis NI, Pallikaris IG,Jankov MR. LASIK flap characteristics using the Moria M2microkeratome with the 90-microm single use head. J RefractSurg 2007;23(1):45–49.

9. Chang JS. Complications of sub-Bowman’s keratomileusiswith a femtosecond laser in 3009 eyes. J Refract Surg
2008;24(1):S97–S101.

1

1

1

2

2

2

10. Reinstein DZ, Archer TJ, Gobbe M, Silverman RH, Cole-man DJ. Epithelial thickness in the normal cornea: three-dimensional display with Artemis very high-frequency digitalultrasound. J Refract Surg 2008;24(6):571–581.

11. Hernandez-Matamoros J, Iradier MT, Moreno E. Treatingfolds and striae after laser in situ keratomileusis. J CataractRefract Surg 2001;27(3):350–352.

12. Sridhar MS, Rao SK, Vajpayee RB, Aasuri MK, Hannush S,Sinha R. Complications of laser-in-situ-keratomileusis. In-dian J Ophthalmol 2002;50(4):265–282.

13. Prakash G, Agarwal A, Yadav A, et al. A prospectiverandomized comparison of four femtosecond LASIK flapthicknesses. J Refract Surg 2010;26(6):392–402.

14. Reinstein DZ, Waring GO. Graphic reporting of outcomes ofrefractive surgery. J Refract Surg 2009;25(11):975–978.

15. Taneri S, Feit R, Azar DT. Safety, efficacy, and stabilityindices of LASEK correction in moderate myopia and astig-matism. J Cataract Refract Surg 2004;30(10):2130–2137.

16. Azar DT, Ghanem RC, de la Cruz J, et al. Thin-flap(sub-Bowman’s keratomileusis) versus thick-flap laser in situkeratomileusis for moderate to high myopia: case-control
analysis. J Cataract Refract Surg 2008;34(12):2073–2078.

7. Moller-Pedersen T, Li H, Petroll WM, Cavanagh HD, JesterJV. Confocal microscopic characterization of wound repairafter photorefractive keratectomy. Invest Ophthalmol VisSci 1998;39(3):487–501.

8. McCulley JP, Petroll WM. Quantitative assessment of cor-neal wound healing following IntraLASIK using in vivoconfocal microscopy. Trans Am Ophthalmol Soc 2008;106:84–90.

9. Rocha KM, Kagan R, Smith SD, Krueger RR. Thresholds forinterface haze formation after thin-flap femtosecond laser insitu keratomileusis for myopia. Am J Ophthalmol 2009;147(6):966–972.

0. Kymionis GD, Karavitaki AE, Portaliou DM, et al. Interfacehaze formation after ultra thin flap laser in situ keratomileu-sis. Ophthalmic Surg Lasers Imaging 2010. Forthcoming.

1. Chang JSM. Refractive surgical problem: January consulta-tion #5. J Cataract Refract Surg 2010;37(1):214–215.

2. Kymionis GD, Portaliou DM, Tsiklis NS, PanagopoulouSI, Pallikaris IG. Thin LASIK flap creation using theSCHWIND Carriazo-Pendular microkeratome. J Refract
Surg 2009;25(1):33–36.

Biosketch

Amar Agarwal is the director of Dr Agarwal’s group of Eye hospitals, Chennai, India. He has to his credit various surgicalinnovations, like Phakonit, Microphakonit, no-anesthesia cataract surgery, gas-forced infusion, glued intraocular lens. Hehas authored more than 40 books in ophthalmology. He is an active member of American Academy of Ophthalmology,and a recipient of Barraquer, Kelman, Casebeer, and American academy’s achievement award. His current researchinterest includes �Aberropia= and Multifocal glued intraocular lens.

COMPARATIVE OUTCOMES OF FEMTOSECOND SUB-BOWMAN KERATOMILEUSISVOL. 152, NO. 4 590.e1

Biosketch

Gaurav Prakash, MD, is Consultant in-charge, Refractive Surgery Services at Dr Agarwal’s group of eye hospitals,Chennai, India. He has done his ophthalmic residency and corneal fellowship from All India Institute of MedicalSciences, New Delhi, India. His research interests span wavefront optics, refractive surgery, corneal disorders and novelinstruments, on which he has authored more than 45 peer-reviewed publications and multiple meeting presentations. Hewishes to pursue a career as a clinician-researcher and contribute to international ophthalmology.

AMERICAN JOURNAL OF OPHTHALMOLOGY590.e2 OCTOBER 2011

Femtosecond Sub-Bowman Keratomileusis: A Prospective, Long-term, Intereye Comparison of Safety and...

Documents

Transcript of Femtosecond Sub-Bowman Keratomileusis: A Prospective, Long-term, Intereye Comparison of Safety and...