Laser in situ keratomileusis buttonhole: Classification and management algorithm

ARTICLE

Laser in situ keratomileusis buttonhole: Classification and management algorithm Mona Harissi-Dagher, MD, Amit Todani, MD, Samir A. Melki, MD, PhD

PURPOSE: To report the classification, management, and visual outcomes after laser in situ keratomileusis (LASIK) flap buttonhole caused by a microkeratome cut. SETTING: Private practice, Boston, Massachusetts, USA. METHODS: This retrospective observational case series comprised 15 patients with an intraoperative LASIK flap buttonhole or near buttonhole. In all cases, the flap was left in place or repositioned without excimer laser treatment. Buttonholes were classified by stage, and a treatment algorithm based on the stage was devised to determine the timing and type of intervention. The uncorrected visual acuity (UCVA), best spectacle-corrected visual acuity (BSCVA), and complications associated with the laser vision correction surgery were reported. RESULTS: Postoperative follow-up ranged from 1 week to 23 months. All 9 patients who were retreated had a postoperative UCVA of 20/25 or better. No retreated patient lost BSCVA. Before retreatment, the median UCVA was 20/80 (range 20/40 1 to counting fingers), the median BSCVA was 20/20 2 (range 20/15 1 to 20/70), and the spherical equivalent (SE) refractive errors ranged from 1.00 to 6.62 diopters (D). After retreatment, the median UCVA was 20/20 2 (range 20/15 1 to 20/25 1), the median BSCVA was 20/20 (range 20/15 to 20/20 3), and the SE refractive errors ranged from C0.50 to 0.75 D. Complications after laser correction treatment included overcorrection in 3 patients and corneal haze in 2 patients. CONCLUSIONS: Classification of buttonholes was helpful in guiding treatment. Good UCVA and BSCVA were achieved by following a simple treatment algorithm based on surface ablation. J Cataract Refract Surg 2008; 34:1892–1899 Q 2008 ASCRS and ESCRS

Laser in situ keratomileusis (LASIK) is a well-described surgical technique designed to reshape the corneal stroma by an excimer laser after creation of a corneal flap. The buttonhole is an intraoperative

Accepted for publication August 6, 2008. From the Massachusetts Eye and Ear Infirmary (Harissi-Dagher, Todani, Melki), Harvard Medical School, and a private practice (Melki), Boston, Massachusetts, USA; Universite´ de Montre´al (Harissi-Dagher), Montreal, Quebec, Canada. No author has a financial or proprietary interest in any material or method mentioned. Presented in part at the annual meeting of the American Academy of Ophthalmology, New Orleans, Louisiana, USA, October 2004. Corresponding author: Samir A. Melki, MD, PhD, Cornea and Refractive Surgery Services, Massachusetts Eye and Ear Infirmary, 243 Charles Street, Boston, Massachusetts 02114, USA. E-mail: [email protected].

1892

Q 2008 ASCRS and ESCRS Published by Elsevier Inc.

LASIK complication resulting from an uncut portion of the corneal flap. It may lead to irregular astigmatism and loss of vision.1 The LASIK buttonhole occurs when the microkeratome blade travels more superficially than intended and enters the epithelium–Bowman layer complex. Buttonholes can be full or partial thickness depending on whether the blade exits the overlying epithelium. Partial-thickness buttonholes, also called occult or near buttonholes, describe situations in which the flap defect does not extend above Bowman layer.1 The following is a single-surgeon case series of 15 patients who had a buttonhole or near buttonhole during LASIK over a 6-year period. We describe an algorithm that is based on the clinical stage of buttonholes. The algorithm is designed to streamline the management approach to this adverse event during LASIK. We detail a surface treatment approach based on each stage. Finally, we report the results of the management algorithm. 0886-3350/08/$dsee front matter doi:10.1016/j.jcrs.2008.08.013

1893

LASIK BUTTONHOLE ALGORITHM

Table 1. Patient preoperative data and effect of buttonhole formation on manifest refraction. Pt

Age (Y)/Sex

Date of Surgery

Microkeratome Head

Depth Footplate (mm)

Avg K (D)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

40/F 39/M 30/M 51/M 34/M 41/F 31/F 42/M 27/F 55/M 52/F 32/F 53/M 33/M 29/M

6/2/2006 1/3/2003 3/29/2002 5/17/2002 1/18/2002 12/7/2001 5/24/2006 4/1/2005 4/1/2005 10/24/2003 9/8/2002 8/11/2006 1/28/2005 8/6/2002 9/1/2006

Moria 90 Hansatome 9.5 Hansatome 9.5 Hansatome 9.5 Hansatome 8.5 Hansatome 9.5 Moria 90 Hansatome 9.5 Hansatome 9.5 Hansatome 9.5 Hansatome 9.5 Moria 90 Hansatome 9.5 Hansatome 9.5 Moria 90

90 180 180 160 180 180 90 180 160 180 180 90 180 200 90

43.50 45.30 44.87 45.75 45.12 45.12 46.37 45.00 45.37 45.75 45.25 43.74 43.12 46.62 44.06

Avg K Z average keratometry; BH Z buttonhole; MR Z manifest refraction; NBHZ near buttonhole; Preop Pachy Z preoperative pachymetry; Pt Z patient; SE Z spherical equivalent; VA Z visual acuity

PATIENTS AND METHODS A review of all medical records of patients who had a LASIK flap buttonhole or occult buttonhole between 2001 and the first half of 2007 was performed. All patients had LASIK by the same surgeon (S.A.M.). The surgical technique was identical in all cases. An aseptic field was prepared, and the lids were draped. A wire lid speculum was placed after application of topical anesthesia (proparacaine hydrochloride ophthalmic solution USP 0.5%). Gentian violet fiducial marks were applied to the corneal surface. A Hansatome (Bausch & Lomb) or Moria M2 (Moria) suction ring was applied with a slight superior decentration. Vacuum was applied, and the intraocular pressure (IOP) was checked with a pneumotonometer (Mentor O & O, Inc.) until a minimum reading of 80 mm Hg was attained. Before the IOP was measured, the ring was gently lifted and rotated to ensure against the presence of pseudosuction. The keratome head was then engaged in position, and a forward pass was activated. If a buttonhole was noted, the flap was not lifted unless the edges of the hole were not well apposed. In the latter situation, balanced salt solution through an irrigation cannula was used to refloat the flap. No excimer laser treatment was performed in buttonhole case, and a bandage contact lens was placed until the next morning. In cases in which the buttonhole occurred in the first eye of a planned bilateral simultaneous LASIK, second-eye surgery was delayed to a later date. A buttonhole flap was classified by stage, and a management algorithm was devised to streamline the treatment approach and guide therapy. Postoperatively, retreatment was performed a mean of 12.1 weeks after the initial procedure unless an epithelial ingrowth was noted. Further surgical intervention was based on surface ablation; the LASIK flap was not recut in any case.

RESULTS Tables 1 and 2 show the preoperative and postoperative data. Fifteen buttonholes or near buttonholes occurred,

for a calculated incidence of 0.57%. There were 11 cases of buttonholes (0.41%) and 4 cases of near buttonhole (0.15%). The incidence of buttonholes and near buttonholes was 0.59% and 0.50% for the Hansatome keratome and the Moria M2 keratome, respectively. Nine of the patients were men. The mean age was 39 years (range 27 to 55 years). The mean keratometry was 45.0 diopters (D), and the mean central pachymetry was 552 mm. In 10 of 15 cases, the keratome footplate was 180 mm or thicker. Refraction before retreatment showed a spherical equivalent (SE) shift (mean 0.79 G 1.14 D) toward hyperopia in 13 eyes and toward myopia (mean 0.62 G 0.01 D) in 2 eyes. All buttonholes occurring with the Hansatome keratome had a triangular shape, with the apex of the triangle facing the surgeon (Figure 1, A to C). This was not noted in the 3 patients in whom the Moria keratome was used (Figure 1, D). Follow-up ranged from 1 week to 24 months. Six patients (40%) opted not to have laser vision correction after the buttonhole complication. One patient (patient 9, Table 1) lost 1 line of best spectacle-corrected visual acuity (BSCVA) at the time of the last follow-up visit at 3.5 months. Seven patients had phototherapeutic keratectomy (PTK) followed by photorefractive keratectomy (PRK) and application of mitomycin-C (MMC) 0.02% at the same setting after a minimum waiting period of 12 weeks. Most patients had full resolution of the buttonhole without residual scarring (Figure 2). Two other patients had retreatment earlier than 12 weeks, as detailed below. Of the 9 patients who were retreated, the median uncorrected visual acuity

J CATARACT REFRACT SURG - VOL 34, NOVEMBER 2008

1894

LASIK BUTTONHOLE ALGORITHM

Table 1 (Cont.) Preop Pachy

BH Vs NBH

Triangular?

MR Before BH (VA)

MR After BH (VA)

SE Affected by BH

566.00 588.00 570.00 515.00 563.00 544.00 546.00 530.00 532.00 578.00 556.00 538.00 579.00 550.00 529.00

BH BH BH BH BH BH BH BH BH BH BH NBH NBH NBH NBH

No Yes Yes Yes Yes Yes No Yes Yes Yes Yes No Yes Yes No

1.25 1.00  90 (20/30C1) 2.00 2.50  60 (20/15 2) 5.25 1.25  150 (20/15) 5.00 1.50  90 (20/20C2) 2.25 0.75  120 (20/20C) 4.25 0.75  075 (20/20) 2.00 0.75  85 (20/25 2) 3.00 0.50  110 (20/20) 4.50 0.50  150 (20/20) C1.25 sph (20/25 ) 5.25 sph (20/20) 3.00 sph (20/20) 6.00 1.75  173 (20/20) 5.00 2.25  180 (20/15 2) 7.75 0.25  95 (20/20)

0.50 1.00  90 (20/25 2) 2.75 2.25  65 (20/70) 5.75 1.50  135 (20/20C2) 4.25 1.50  90 (20/20) 1.75 1.25  120 (20/15 1) 4.00 sph (20/25 1) 1.25 sph (20/20 1) 0.50 0.75  115 (20/20 2) 4.00 0.75  45 (20/25C2) C2.25 0.50  35 (20/20) 4.25 0.25  155 (20/20 2) 2.75 sph (20/20 1) 5.25 2.75  175 (20/20 2) 3.50 0.75  180 (20/20 1) 6.75 0.75  150 (20/20)

0.75 D/42.85% plus 0.62 D/19.07% minus 0.63 D/10.73% minus 0.75 D/ 13.04% plus 0.25D/9.54% plus 0.62 D/13.41% plus 1.12 D/47.25% plus 2.38 D/73.23% plus 0.38 D/08.00% plus 1.25D/100.00% plus 0.88 D/39.11% plus 0.25D/8.33% plus 0.25D/3.63% plus 2.25 D/36.76% plus 0.75 D/9.52% plus

(UCVA) before retreatment was 20/80 (range 20/40 1 to counting fingers), the median BSCVA was 20/20 2 (range 20/15 1 to 20/70), and the SE refractive errors ranged from 1.00 to 6.62 D. After retreatment, the median UCVA was 20/20 2 (range 20/15 1 to 20/ 25 1), the median BSCVA was 20/20 (range 20/15 to 20/20 3), and the SE refractive errors ranged from C0.50 to 0.75 D. One patient (patient 12, Tables 1 and 2) had retreatment 7 weeks after the buttonhole because she showed stability of refraction and insisted on quicker visual rehabilitation. One patient (patient 14, Tables 1 and 2) had PTK followed by PRK with MMC 10 days after

the buttonhole due to the presence of an epithelial ingrowth at the edge of the buttonhole. This patient had an early overcorrection that resolved over time. Transient corneal haze and overcorrection occurred in 2 patients (patients 13 and 14, Tables 1 and 2) who had PTK followed by PRK with MMC for retreatment of a near buttonhole. The corneal haze and overcorrection resolved after topical prednisolone treatment. Patient 6 had overcorrection not associated with haze (C4.00 –0.75  145) after PTK followed by PRK with MMC for buttonhole retreatment (Table 2). Twenty-four weeks after the latter treatment, the epithelium was gently removed after application of 20% alcohol in

Table 2. Management and results after LASIK flap buttonhole.

Pt

Treatment

Treatment Timing After BH (Wk)

Ultimate UCVA

1 2 3

PTK/PRK 50 mm PTK/PRK 50 mm PTK/PRK 63 mm

16 13 15

20/25 20/25 20/20

4

PTK/PRK 51 mm

12

20/20

5 6 12

PTK/PRK 69 mm PTK/PRK 62 mm PTK/PRK 50 mm

21 12 7

20/15 1 20/25 20/20C2

13 14

PTK/PRK 51 mm PTK/PRK 52 mm

13 10 (d)

20/20 20/20

1

2

3 1

Ultimate MR 0.50 0.50  80 Plano 1.25  60 Not done as VAR20/20 Not done as VAR20/20 0.50 sph Plano 0.50  50 Not done as VAR20/20 Plano 0.75  165 C0.50 sph

Ultimate BSCVA 20/20 20/20 20/20

3

20/20

2

1

20/15 20/20 20/20C2 20/20 20/20

1

Complications After Treatment

Last FU (Mo)

None None None

15 14 14

None

9

None Overcorrection† None

10 11 8

Haze, overcorrection†† Haze, overcorrection††

23 15

BH Z buttonhole; BSCVA Z best spectacle-corrected visual acuity; FU Z follow-up; MR Z manifest refraction; NBH Z near buttonhole; PRKZ photorefractive keratectomy; Pt Z patient; PTKZ phototherapeutic keratectomy; UCVAZ uncorrected visual acuity; VA Z visual acuity † Overcorrection required further intervention with hyperopic laser-assisted subepithelial keratectomy. †† Overcorrection gradually resolved over time without further surgical intervention.

J CATARACT REFRACT SURG - VOL 34, NOVEMBER 2008

LASIK BUTTONHOLE ALGORITHM

1895

Figure 1. Stage 1 triangular buttonholes in LASIK flaps created by the Hansatome keratome (A, B, and C). Buttonholes in flaps created by the Moria M2 keratome were not triangular (D). Staining with sodium fluorescein (0.25%) allowed close observation of epithelial cell migration under the buttonhole edges through a fistula.

a metallic well for 60 seconds followed by a hyperopic ablation. A contact lens was placed until full reepithelialization. The final refraction was plano 0.50  50 with a BSCVA of 20/20. DISCUSSION The incidence of LASIK buttonholes has been reported to vary between 0.20% and 0.56%.2–4 This case series represents an incidence of 0.57%, which is in agreement with previously published figures. The advent of the femtosecond laser flap creation may have reduced the risk of this complication during LASIK. Nevertheless, vertical bubble breakthrough with the femtosecond laser was recently reported in the peerreviewed literature.5

Several theories have been presented to explain the occurrence of LASIK buttonholes due to microkeratome cuts. Buttonholes occur when an area in the cornea is missed by the blade and remains uncut. The microkeratome blade exits the epithelium prematurely as it travels across the stroma and quickly reenters to follow the original path.6 Steep corneas and lack of synchronization between the translational keratome movement and oscillatory blade movement have been suggested as causes of buttonholes.2,7 Pulaski8 postulates that a partially opened eye can lead to dessication and thinning of the central cornea, which may result in a buttonhole during the keratome pass. Poor blade quality has also been suggested as a cause, especially when a cluster of cases is encountered.4

Figure 2. Buttonhole before (A) and after retreatment using PTK followed by PRK with MMC (B). A small epithelial ingrowth was noted at the inferior edge of the buttonhole (Stage 2). This did not warrant early intervention as it resolved spontaneously.

J CATARACT REFRACT SURG - VOL 34, NOVEMBER 2008

1896

LASIK BUTTONHOLE ALGORITHM

In our case series, the mean keratometry reading was 45.0 D. This is not consistent with steep curvature as a common etiology.2 A recent study by AlbeldaValle´s et al.9 also found no relationship between steep keratometry and buttonhole occurrence. Nevertheless, one has to consider possible selection bias as patients with keratometric values higher than 47.5 D typically have surface ablation at our center. From all presumed etiologies, we suspect microsuction loss to be the main reason behind flap buttonholes in our case series. Inadequate suction can occur in cases of conjunctival incarceration in the suction port (pseudosuction) or simply from an inadequately functioning vacuum unit.10 On occasion, high astigmatism compromises the adherence of the suction ring to the cornea.11 We have noticed that LASIK buttonholes with the Hansatome keratome are invariably triangular (Figure 1, A to C) (S. Melki, MD, et al, ‘‘The LASIK Buttonhole or the Bermuda Triangle,’’ video presented at ASCRS Symposium on Cataract, IOL and Refractive Surgery, San Diego, California, USA, May 2004). Typically, the apex of the triangle faces the surgeon. The presence of this triangle sign on the surface of the cornea after the keratome pass should alert the surgeon to carefully inspect the flap before considering lifting it. It may also be helpful when examining patients with visual complaints after LASIK when access to their surgical record is not available. Buttonholes with another superior hinged keratome (Moria M2) do not seem to be triangular (Figure 1, D). The difference in buttonhole shape between these 2 superior hinged keratomes is unclear. If a buttonhole is encountered, it is best to cancel the procedure, rather than lift the flap or reposition it if it had to be lifted.12 Laser treatment should not

be performed at that time.1 The flap should be carefully inspected at the slitlamp to ensure good apposition of the buttonhole margins. We also advise aborting the procedure in cases of near buttonholes. In these cases, Bowman layer is found uncut in the central or peripheral cornea; however, the epithelium remains grossly intact. Despite the latter, epithelial ingrowth can occur and complicate the postoperative course.13–15 A patient referred to us with visual complaints after LASIK had an epithelial ingrowth after laser treatment was initiated, despite the presence of a near buttonhole (Figure 3, A). This resulted in irregular astigmatism (Figure 3, B) and was a challenge to treat. Recutting a flap once a buttonhole has occurred is not advisable due to the risk of double flaps, loss of tissue, and subsequent irregular astigmatism.15,16 Similarly, we avoid recutting with the femtosecond laser due to the manipulation required to lift the flap as well as the risk of margin disruption by cavitation bubbles. By observing the natural history of LASIK buttonholes, a classification emerges to direct the management approach. Full- and partial-thickness buttonholes should not be classified differently as they behave in a similar manner and thus require similar management. Buttonholes can occur in the periphery or the center of the corneal flap. Central and peripheral buttonholes also behave similarly (F. Yaghouti, MD, personal communication, January 2006) and should probably not be classified differently other than by location.1 Laser in situ keratomileusis buttonholes are probably best classified in 3 stages. Stage 1 is a full- or partial-thickness buttonhole without epithelial ingrowth (Figure 1). Stage 2 is a full- or partialthickness buttonhole with epithelial ingrowth

Figure 3. A patient with decreased visual acuity after LASIK secondary to epithelial ingrowth infiltrating through a near buttonhole (A). Corneal topography shows an area of irregular astigmatism overlying the epithelial ingrowth (B). This represents a stage 3 near buttonhole. J CATARACT REFRACT SURG - VOL 34, NOVEMBER 2008

LASIK BUTTONHOLE ALGORITHM

(Figure 2, A). Stage 3 is a full- or partial-thickness buttonhole with epithelial ingrowth and resultant stromal melt/scarring or flap elevation. The latter can be detected by slitlamp biomicroscopy (Figure 3, A) or by an area of irregular astigmatism on corneal videokeratography (Figure 3, B). We devised a treatment algorithm based on the 3-stage buttonhole classification (Figure 4). We have found this algorithm to be helpful in ensuring good visual outcomes. The infiltration of epithelial cells through the buttonhole determines which course of action is best. At stage 1, close observation is paramount. The epithelial disruption at the margins of the buttonhole increases the risk for diffuse lamellar keratitis (DLK) and for epithelial ingrowth through the edges of the hole. Hourly topical steroids used in the first 2 to 3 days help minimize the risk for DLK. It is critical to prevent DLK from progressing beyond stage 2 as lifting the flap for irrigation of the inflammatory cells will disrupt the buttonhole edges and may lead to further

1897

difficulties.17 As long as the edges of the buttonhole stain with fluorescein, a high level of suspicion should be maintained for an epithelial fistula and consequent potential cellular infiltration (Figure 1, C and D). It is best to wait at least 12 weeks before deciding on further action after a buttonhole occurs.12 Kapadia et al.18 recommend early intervention before the formation of significant scarring. We believe that a longer period of time allows epithelial hyperplasia to smooth the corneal surface and ensures refractive stability. Our results confirm the possible benefit of a longer waiting period. If slitlamp biomicroscopy reveals a smooth epithelial surface, a 50 mm PTK ablation followed by the refractive treatment is performed. In this situation, the epithelium acts as a masking agent during the PTK–PRK ablation. It allows smoothing of stromal scars and irregularities. The PTK–PRK approach in this situation has been described by us13 and by others.19 Figure 2 shows a LASIK buttonhole before and after treatment by PTK followed by PRK with MMC.

Figure 4. Algorithm depicting the appropriate surgical management by stage after LASIK flap buttonhole occurrence.

J CATARACT REFRACT SURG - VOL 34, NOVEMBER 2008

1898

LASIK BUTTONHOLE ALGORITHM

If after 12 weeks the epithelial surface is deemed more irregular than the underlying buttonholed stroma, it may be best to remove it with 20% alcohol applied for 60 seconds. The alcohol time is critical in avoiding adherence of the epithelium to the edges of the hole, which may disrupt it during peeling. Care should be taken to prevent alcohol contact with the peripheral edges of the flap. Choosing an appropriately sized well and applying a methylcellulose solution to the flap edges should provide adequate protection from the alcohol solution. Although surface ablation on a LASIK flap is reported to be associated with an elevated risk for corneal scarring,12 the advent of MMC as an adjunctive antiscarring agent has permitted a more liberal use of that modality in these situations.20–22 We use 60 seconds of MMC 0.02% application to prevent stromal haze formation.13 We have not observed permanent haze in any of our patients. This is especially important because the considerable haze would be difficult to scrape. In our study, overcorrection occurred in 3 (33%) of the 9 patients who had PTK–PRK treatment. It resolved without intervention in 2 patients. Persistent overcorrection could be related to epithelial irregularity with a multifocal surface after buttonhole formation, leading to erroneous pretreatment refractions. It is also possible that the time needed to switch between a PRK and PTK treatment card leads to stromal dehydration with subsequent overcorrection. It is unlikely that overcorrection was related to the MMC 0.02% as this agent was used in all cases without significant refractive nomogram adjustment. Consideration could be given to reducing the amount of the PTK or the PRK ablation at the time of retreatment until a better understanding of the etiology of overcorrection in these situations is attained. Stage 2 refers to the presence of an epithelial ingrowth around the margins of a buttonhole or near buttonhole (Figure 2, A). A high index of suspicion should be maintained in the early postoperative period. This can be a disturbing complication because stromal melting and irregular astigmatism may follow. If the ingrowth is deemed small and nonprogressive, it can be observed for 12 weeks. It can then be treated similarly to a stage 1 buttonhole (Figure 4). If the ingrowth appears to be progressive and/or threatening stromal melting, immediate intervention is warranted to prevent evolution to stage 3. To avoid the latter, we recommend a staged approach. The goal of the first stage is essentially therapeutic and is intended to obliterate the infiltrated epithelial cells.23 An initial 6.5 mm PTK treatment is applied at a depth of 40 to 50 mm. If a residual epithelial ingrowth is noted on biomicroscopy, 10 mm PTK increments can be

applied with intermittent slitlamp examination until the ingrowth is satisfactorily ablated. The ingrowth was ablated by the initial 50 mm PTK in the only patient in our series who developed a significant ingrowth (patient 14, Tables 1 and 2). If the total amount of laser pulses applied in the first stage exceeds 50 mm, it may be best to delay the refractive treatment to a later stage until a new stable refraction is obtained. At that time, surface ablation can be applied to correct the remaining refractive error. This approach would avoid the risk for a hyperopic shift that would result if the full refractive treatment were applied in addition to a significant PTK ablation. Stage 3 buttonholes (Figure 3) are the most difficult to treat and may result in significant loss of vision. They occur if an epithelial ingrowth is significant enough to cause irregular astigmatism due to elevation of the flap or stromal melting. A hard contact lens over refraction will differentiate between vision loss due to an obstructing scar versus irregular astigmatism. Lifting the flap to clear persistent epithelial ingrowth may result in flap tears if considerable edge scars have formed. A PTK/PRK treatment may not be successful at smoothing the underlying stromal irregularity. Careful consideration of the benefits and risks of each approach and its timing should be individualized to the corresponding situation. Treatment of LASIK buttonholes in hyperopic patients may have to be approached differently. In that situation, most of the laser corrective ablation is in the periphery and may not ablate the stromal scar. Surgical planning must take into consideration the buttonhole scar effect on vision. If the latter is considered significant, this may be best approached with a 2-stage treatment: a primary PTK treatment to ablate the scar followed by a later PRK treatment once refractive stability is achieved. In conclusion, the cause of microkeratome-related LASIK buttonholes is still elusive and preventive measures are therefore difficult to pinpoint. In our series, most cases had average keratometry readings, were mostly fashioned with a deeper keratome footplate, and showed no gross evidence of suction loss during surgery. By following our treatment algorithm, we obtained good results in these patients with buttonholed flaps. The etiology of LASIK buttonholes and the overcorrection after PTK–PRK treatment require further investigation. REFERENCES 1. Melki SA, Azar DT. LASIK complications: etiology, management, and prevention. Surv Ophthalmol 2001; 46:95–116 2. Gimbel HV, Anderson Penno EE, van Westenbrugge JA, Ferensowicz M, Furlong MT. Incidence and management of intraoperative and early postoperative complications in 1000

J CATARACT REFRACT SURG - VOL 34, NOVEMBER 2008

1899

LASIK BUTTONHOLE ALGORITHM

3. 4.

5.

6. 7.

8. 9.

10.

11.

12.

13.

14.

consecutive laser in situ keratomileusis cases. Ophthalmology 1998; 105:1839–1847; discussion by TE Clinch, 1847–1848 Lin RT, Maloney RK. Flap complications associated with lamellar refractive surgery. Am J Ophthalmol 1999; 127:129–136 Stulting RD, Carr JD, Thompson KP, Waring GO III, Wiley WM, Walker JG. Complications of laser in situ keratomileusis for the correction of myopia. Ophthalmology 1999; 106:13–20 Srinivasan S, Herzig S. Sub-epithelial gas breakthrough during femtosecond laser flap creation for LASIK [video report]. Br J Ophthalmol 2007; 91:1373 Wilson SE. LASIK: management of common complications. Cornea 1998; 17:459–467 Leung ATS, Rao SK, Cheng ACK, Yu EWY, Fan DSP, Lam DSC. Pathogenesis and management of laser in situ keratomileusis flap buttonhole. J Cataract Refract Surg 2000; 26:358–362 Pulaski JP. Etiology of buttonhole flaps [letter]. J Cataract Refract Surg 2000; 26:1270–1271 Albelda-Valle´s JC, Martin-Reyes C, Ramos F, Beltran J, Llovet F, Baviera J. Effect of preoperative keratometric power on intraoperative complications in LASIK in 34,099 eyes. J Refract Surg 2007; 23:592–597 Davidorf JM, Zaldivar R, Oscherow S. Results and complications of laser in situ keratomileusis by experienced surgeons. J Refract Surg 1998; 14:114–122 Zaldivar R, Davidorf JM, Oscherow S. Laser in situ keratomileusis for myopia from 5.50 to 11.50 diopters with astigmatism. J Refract Surg 1998; 14:19–25 Lichter H, Stulting RD, Waring GO III, Russell GE, Carr J. Buttonholes during LASIK: etiology and outcome. J Refract Surg 2007; 23:472–476 Taneri S, Koch JM, Melki SA, Azar DT. Mitomycin-C assisted photorefractive keratectomy in the treatment of buttonholed laser in situ keratomileusis flaps associated with epithelial ingrowth. J Cataract Refract Surg 2005; 31:2026–2030 Grupcheva CN, Malik TY, Craig JP, McGhee CNJ. In vivo confocal microscopy of corneal epithelial ingrowth through a laser in situ keratomileusis flap buttonhole. J Cataract Refract Surg 2001; 27:1318–1322

15. Tekwani NH, Chalita MR, Krueger RR. Secondary microkeratome-induced flap interference with the pathway of the primary flap. Ophthalmology 2003; 110:1379–1383 16. Rubinfeld RS, Hardten DR, Donnenfeld ED, Stein RM, Koch DD, Speaker MG, Frucht-Pery J, Kameen AJ, Negvesky GJ. To lift or recut: changing trends in LASIK enhancement. J Cataract Refract Surg 2003; 29:2306–2317 17. Linebarger EJ, Hardten DR, Lindstrom RL. Diffuse lamellar keratitis: diagnosis and management. J Cataract Refract Surg 2000; 26:1072–1077 18. Kapadia MS, Wilson SE. Transepithelial photorefractive keratectomy for treatment of thin flaps or caps after complicated laser in situ keratomileusis. Am J Ophthalmol 1998; 126:827–829 19. Muller LT, Candal EM, Epstein RJ, Dennis RF, Majmudar PA. Transepithelial phototherapeutic keratectomy/photorefractive keratectomy with adjunctive mitomycin-C for complicated LASIK flaps. J Cataract Refract Surg 2005; 31:291–296 20. Lane HA, Swale JA, Majmudar PA. Prophylactic use of mitomycin-C in the management of a buttonholed LASIK flap. J Cataract Refract Surg 2003; 29:390–392 21. Majmudar PA, Forstot SL, Dennis RF, Nirankari VS, Damiano RE, Brenart R, Epstein RJ. Topical mitomycin-C for subepithelial fibrosis after refractive corneal surgery. Ophthalmology 2000; 107:89–94 22. Solomon R, Donnenfeld ED, Perry HD. Photorefractive keratectomy with mitomycin C for the management of a LASIK flap complication following a penetrating keratoplasty. Cornea 2004; 23:403–405 23. Vajpayee RB, Gupta V, Sharma N. PRK for epithelial ingrowth in buttonhole after LASIK. Cornea 2003; 22:259–261

First author: Mona Harissi-Dagher, MD Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston, Massachusetts, USA

J CATARACT REFRACT SURG - VOL 34, NOVEMBER 2008