Wound Dehiscence in a Patient With Keratoconus After Penetrating Keratoplastyand LASIK

Residual refractive error often complicates penetrating keratoplasty(PKP) and can be treated surgically in patients who are contact lens intolerant.Both photorefractive keratectomy and laser in situ keratomileusis (LASIK)have been used to reduce postoperative myopia and astigmatism following PKP.¹ Reported complications of LASIK following PKP includeirregular astigmatism, photoablation decentration, stromal bed hemorrhage,obstruction of the microkeratome path by graft sutures, corneal perforation,flap dislocation, melts, slippage, paracentral perforation, or buttonhole.Keratoconus has been described as a contraindication to LASIK due to cornealinstability attributed to the exceedingly thin residual stromal bed producedby the lamellar dissection,² but severalstudies have documented successful LASIK following PKP in patients with keratoconus.^3-5 We report hereina case of wound dehiscence following LASIK in a patient with post-PKP keratoconus,the first such report to our knowledge.

A 59-year-old man with a history of keratoconus, PKP, and LASIK in hisright eye was seen in our service with wound dehiscence following LASIK. Thepatient first underwent PKP in both eyes 30 years earlier for keratoconus,followed by photorefractive keratectomy in both eyes for postoperative astigmatism6 years ago. However, the postoperative course in the right eye was complicatedby the development of corneal haze, topographic irregularity, and an earlycataract that reduced the best rigid lens–corrected visual acuity to20/50 OD and produced disabling symptoms of glare and light scatter. Threeyears ago the patient underwent repeated PKP, cataract extraction, and intraocularlens placement in his right eye. Visual acuity improved postoperatively to20/25 OD with a rigid contact lens. However, because of contact lens intolerance,the patient elected to undergo LASIK. A microkeratome (Hansatome; Bausch &Lomb, Rochester, NY) was used to create the LASIK flap, with the goal of creatinga flap that was only slightly larger than the 8.0-mm-diameter corneal graft(Figure 1). The operating surgeonreported that the flap appeared to be thin centrally but that a full-thicknessbuttonhole was not evident. The stromal bed was inspected and noted to besmooth and regular, so the ablation was completed as planned, and the stromalflap was repositioned without difficulty.

On the first postoperative day, the inferior aspect of the flap wasnoted to be unusually edematous and there was relative thinning of the flapat the corneal apex, but no other abnormalities including a shallow chamber,interface fluid, or dehiscence were identified. Two weeks after the LASIKprocedure, the patient reported irritation of the right eye that compelledhim to rub the eye with moderate vigor. He noted excessive tearing, and soon,thereafter, experienced a progressive decrease in visual acuity and pain.He was seen by a local ophthalmologist who identified wound dehiscence anda flat anterior chamber and referred the patient to our service. On our examination,the visual acuity was 20/400 OD, and the eye was hypotonous with an intraocularpressure of 0 mm Hg. Slitlamp biomicroscopic examination revealed a flat anteriorchamber and a positive Seidel test at the 6-o'clock position. Fluoresceinwas noted to stream out from beneath the flap edge, but the overlying flapwas edematous, and a dehiscence in the underlying and more centrally locatedgraft-host junction could not be clearly identified. Inspection of the centralcornea revealed a relatively lucent zone indicative of relative flap thinningconsistent with a partial-thickness buttonhole (Figure 2).

Four interrupted 10.0 nylon sutures were placed inferiorly, traversingthe graft-host junction and LASIK flap edge. The next day, visual acuity remained20/400 OD, but the anterior chamber appeared deep and formed, the Seidel testresult was negative, and the intraocular pressure was 13.5 mm Hg. An areaof uplift, edema, and necrosis of the LASIK flap appeared along the nasaledge of the graft and LASIK flap (Figure 3), and further sutures were placed to correct the uplift. Two weeksfollowing placement of these sutures, visual acuity remained at 20/400 ODowing to high irregular suture-induced astigmatism, but the flap edema hadresolved, the flap edge necrosis and thinning appeared to have been arrested,there was good flap edge surface contour, there was no evidence of epithelialingrowth, and the anterior chamber was deep and formed.

Although successful LASIK procedures have been reported in patientswith keratoconus following PKP, it would seem obvious that the reduced stromalbed thickness of the keratoconic host would pose a risk for wound dehiscence.However, based on a search of the PubMed database, our description appearsto represent the first reported case of wound dehiscence in this patient group.In patients with keratoconus who have undergone full-thickness grafting, thegraft-host junction consists of thinned host stroma fused to a normal-thicknessgraft. This junction is particularly thin inferiorly because of the inferiordisplacement of the cone in patients with keratoconus. A LASIK flap createdlarger than the graft requires the flap to cross the graft-host junction intothe thinned keratoconic host. As a result, the already limited adherence betweenhost and graft is reduced by the thickness of the flap (Figure 4), thereby increasing the risk of dehiscence either at thetime of surgery, or as appears to be the case described herein, followingminor trauma. Surgeons and patients who consider a LASIK procedure in thissetting should note this increased risk. Adequate time following PKP shouldbe allowed to maximize healing of the graft-host junction, and the wound shouldbe carefully evaluated preoperatively to identify particularly thin regionsthat are at increased risk for dehiscence. If LASIK is performed, the riskof graft dehiscence is likely to be reduced by limiting the diameter of theflap to within the graft-host junction, or creating a thinner flap that leavesa greater area of posterior graft-host adherence. However, as demonstratedin our case, the risk of buttonhole formation appears to be elevated whenLASIK is performed following PKP; therefore, we would recommend that if appropriate,surgeons consider photorefractive keratectomy correction, thereby avoidingthe risk of dehiscence or buttonhole formation.

The authors have no relevant financial interest in this article.

Corresponding author and reprints: Stephen D. McLeod, MD, Universityof California San Francisco, Department of Ophthalmology, 10 Koret Way, K-301,San Francisco, CA 94143 (e-mail: smcleod@itsa.ucsf.edu).