Clinicopathologic Reports, Case Reports, and Small Case Series
June 2003

Pocket of Fluid in the Lamellar Interface After Penetrating Keratoplasty and Laser In Situ Keratomileusis

Author Affiliations



Copyright 2003 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2003

Arch Ophthalmol. 2003;121(6):894-896. doi:10.1001/archopht.121.6.894

A pocket of fluid developing in the lamellar interface is a new complication that has previously been described in patients with steroid-induced intraocular hypertension after laser in situ keratomileusis (LASIK). The first case of interface fluid collection following LASIK was reported in 1999. Steroid-induced intraocular pressure (IOP) elevation and epithelial ingrowth were notable features in this case.1 Subsequently, more cases of interface fluid and steroid-induced elevation of IOP have been reported.25 The pathophysiologic mechanism behind the fluid collection in these cases is presumably the high IOP, which causes diffusion of aqueous humor across the endothelium into the interface, creating a fluid pocket in the lamellar interface. Clinically, this is usually associated with a fluid pocket visible on slitlamp examination in the lamellar interface, decreased visual acuity, a myopic shift in refraction, the presence of microcystic epithelial edema peripheral to the lamellar flap, an increase in central corneal thickness, steepening of corneal topography, and paradoxical hypotony when measuring IOP over the area of the fluid. It usually resolves by treating the high IOP. We report a case of lamellar interface fluid caused by graft failure from endothelial decompensation, without any evidence of intraocular hypertension. To our knowledge, this is the first histopathologic demonstration of a pocket of fluid in the lamellar interface and the first such case caused by endothelial decompensation.

Report of a Case

A 68-year-old man with a history of trauma to the right eye resulting in corneal edema required corneal transplantation. Three years after the penetrating keratoplasty, his uncorrected visual acuity was 20/200 OD, and his best-corrected visual acuity was 20/80 OD with correction of his myopia and high astigmatism. Corneal topography showed irregular astigmatism, with keratometry readings of 36.4 × 48.8 at 15°. As he was unable to wear a contact lens successfully, he subsequently underwent LASIK and astigmatic keratotomy, which was performed to reduce the refractive error of –5.50 + 7.50 × 004. Attempted preoperative endothelial cell density showed extremely low endothelial cell density with large variation in endothelial size and shape. Preoperative pachymetry was 550 mm, and IOP was 13 mmHg. Despite an uncomplicated surgical course, he developed a pocket of fluid in the lamellar interface 3 months postoperatively(Figure 1 and Figure 2), and his uncorrected visual acuity and best-corrected visual acuity deteriorated to counting fingers. His IOP by applanation was 0 mmHg over the cyst and 15 mmHg outside the cyst by applanation tonometry. He subsequently required repeated penetrating keratoplasty. The replaced corneal button was submitted for histopathologic analysis, which showed an extremely reduced number of endothelial cells and stromal edema, with an interface space present where the keratome-created lamellar scar had separated. During permanent section processing, the tissue was dehydrated, thereby removing the fluid in the interface space and leaving the empty space seen on our photomicrograph(Figure 3).

Figure 1.
Image not available

Slitlamp photograph with diffuse illumination of a corneal graft, with a pocket of fluid in the lamellar interface.

Figure 2.
Image not available

Slitlamp photograph with slitbeam illumination of a corneal graft with a pocket of fluid in the lamellar interface.

Figure 3.
Image not available

Photomicrograph of a section of the corneal button, demonstrating reduced endothelial cell numbers and an empty lamellar interface pocket (periodic acid–Schiff; original magnification×20).


This case is unusual because the interface fluid pocket was caused by primary endothelial decompensation, which is supported by the endothelial cell counts and the histopathologic examination of the cornea (Figure 3). We could find no similar cases in a MEDLINE review of the literature. The normal IOP measured outside the LASIK flap supports endothelial decompensation as the primary cause of the interface fluid accumulation. This also appears to be the first histopathologic demonstration of a pocket of fluid in the lamellar interface (Figure 4 and Figure 5). For comparison, we have included histopathologic findings in a normal post-LASIK cornea (Figure 6) and a cornea with endothelial decompensation(Figure 7). In patients who have not undergone LASIK, endothelial decompensation is shown histopatholgically as fluid that collects intracellularly in the basal layer of the epithelium and/or diffusely in the corneal stroma. In this case, the lamellar incision predisposes this site to fluid collection.

Figure 4.
Image not available

Photomicrograph of a section of the lamellar interface pocket (hematoxylin-eosin; original magnification ×100).

Figure 5.
Image not available

Photomicrograph of a section of the lamellar interface pocket (periodic acid–Schiff; original magnification×100).

Figure 6.
Image not available

Photomicrograph of a section of a normal cornea that had undergone laser in situ keratomileusis (periodic acid–Schiff; original magnification ×100).

Figure 7.
Image not available

Photomicrograph of a section of a cornea with pseudophakic bullous keratopathy (periodic acid–Schiff; original magnification ×100).

It is important to note the potential for interface fluid collection as a result of endothelial decompensation in patients undergoing LASIK. Because most previous reports indicate that no significant endothelial damage occurs with the standard LASIK practice of leaving at least a residual bed of 250µm, 6,7 the development of a pocket of fluid in the lamellar interface in this case is most likely due to the large loss of endothelial cells that occurs in eyes that have had penetrating keratoplasty. Long-term studies have shown that eyes that have undergone uncomplicated penetrating keratoplasty have an average annual endothelial cell loss rate of 17.3% for the first 3 years postoperatively, 7.8% from 3 to 5 years, and 2.5% from 5 years or longer8 compared with normal eyes, which have an average annual endothelial cell loss rate estimated at 0.3% to 1% per year.810 We speculate that this large amount of primary endothelial cell loss is the reason this complication has thus far only been seen in eyes following penetrating keratoplasty. Many of these eyes that have undergone corneal transplantation do not show the typical guttata that would alert the surgeon to the possibility of endothelial cell loss, so specular microscopy is especially valuable in detecting this complication. It seems likely that more of these and similar cases will be observed after LASIK, not only in patients who have had keratoplasty, but possibly also in patients who have corneal endothelial dystrophies or who experience future surgical trauma. Additionally, the length of time during which the lamellar interface is vulnerable to fluid collection is unclear.

Back to top
Article Information

Corresponding author and reprints: Daniel M. Albert, MD, MS, F4/338 Clinical Science Center, 600 Highland Ave, Madison, WI 53792-3220.

Lyle  WAJin  GJC Interface fluid associated with diffuse lamellar keratitis and epithelial ingrowth after laser in situ keratomileusis. J Cataract Refract Surg. 1999;251009- 1012Article
Najman-Vainer  JSmith  RJMaloney  RK Interface fluid after LASIK: misleading tonometry can lead to end-stage glaucoma [letter]. J Cataract Refract Surg. 2000;26471- 472Article
Rehany  UBersudsky  VRumelt  S Paradoxical hypotony after laser in situ keratomileusis. J Cataract Refract Surg. 2000;261823- 1826Article
Portellinha  WKuchenbuk  MNakano  KOliveira  M Interface fluid and diffuse corneal edema after laser in situ keratomileusis. J Refract Surg. 2001;17S192- S195
Hamilton  DRManche  EERich  LFMaloney  RK Steroid-induced glaucoma after laser in situ keratomileusis associated with interface fluid. Ophthalmology. 2002;109659- 665Article
Collins  MJCarr  JDStulting  RD  et al.  Effect of LASIK on the corneal endothelium 3 years postoperatively. Am J Ophthalmol. 2001;1311- 6Article
Marshall  JTrokel  SRothery  SSchubert  H An ultrastructural study of corneal incisions induced by an excimer laser at 193-nm. Ophthalmology. 1985;92749- 758Article
Bourne  WMHodge  DONelson  LR Corneal endothelium five years after transplantation. Am J Ophthalmol. 1994;118185- 196
Yee  RWMatsuda  MSchultz  ROEdelhauser  HF Changes in the normal corneal endothelial cellular pattern as a function of age. Curr Eye Res. 1985;4671Article
Cheng  HJacobs  PMMcPherson  KNoble  MJ Precision of cell density estimates and endothelial cell loss with age. Arch Ophthalmol. 1985;1031478Article