DUA'LASER

Human Corneal Anatomy Redeneder

Sc,

es

orr weduwendResults: Three types of BB were obtained. Type-1, is a well-circumscribed, central domeeshapedelevation up to 8.5 mm in diameter (n 14). Type-2, is a thin-walled, large BB of maximum 10.5 mmdiameter, which always started at the periphery, enlarging centrally to form a large BB (n 5), and a mixedtype (n 3). With type-1 BB, unlike type-2 BB, it was possible to peel off DM completely without deating theBB, indicating the presence of an additional layer of tissue. A type-1 BB could be created after rst peelingoff the DM (n 5), conrming that DM was not essential to create a type-1 BB. The popping pressure was1.45 bar and 0.6 bar for type-1 BB and type-2 BB, respectively. Histology conrmed that the cleavageoccurred beyond the last row of keratocytes. This layer was acellular, measured 10.153.6 micronscomposed of 5 to 8 lamellae of predominantly type-1 collagen bundles arranged in transverse, longitudinal,and oblique directions.

Conclusions: There exists a novel, well-dened, acellular, strong layer in the pre-Descemets cornea. Thisseparates along the last row of keratocytes in most cases performed with the BB technique. Its recognition willhave considerable impact on posterior corneal surgery and the understanding of corneal biomechanics andposterior corneal pathology such as acute hydrops, Descematocele and pre-Descemets dystrophies.

Financial Disclosure(s): The authors have no proprietary or commercial interest in any materials discussedin this article. Ophthalmology 2013;-:-- 2013 by the American Academy of Ophthalmology.

The human cornea is made of the epithelium and Bowmanszone anteriorly and the endothelium with its basementmembrane, the Descemets membrane (DM), posteriorly,sandwiching the stroma. The anterior stroma is morecompact than the posterior and differs in the composition ofproteoglycans.1-3

Contemporary corneal transplantation involves selectivereplacement of the affected layer. Deep anterior lamellarkeratoplasty (DALK) allows replacement of affected stromawhile retaining the recipients healthy DM and endothe-lium,4,5 thus eliminating the risk of endothelial rejection.The DM can be separated from the stroma by injecting air,viscoelastic, or saline.4-9 Collectively, these techniques arereferred to as Descemets baring techniques, where it isclaimed that the cleavage occurs at a plane that enables theDM to be laid bare.8,9 The most popular technique is the big

bubble (BB) technique5 wherein air is injected into thecorneal stroma to affect a separation of the DM from theposterior stroma. Air accumulates between these 2 layersin the form of a BB, hence the name. Based on ourclinical experience with DALK we had hypothesized theexistence of a pre-Descemets posterior stromal layerand presented pilot data at the annual congresses ofSocieta Italiana Cellule Staminalie Supercie Oculare andthe Royal College of Ophthalmologists, UK, in 2007.

In this study, we examined the plane of cleavage ex vivoin human donor whole eyes and sclerocorneal discs byinjecting air into the corneal stroma, as is done during theclinical DALK procedure, and present conclusive data todemonstrate the presence of a well-dened, hithertounknown layer in the human posterior cornea. We havetermed this layer Duas layer (DL).

1 2013 by the American Academy of Ophthalmology ISSN 0161-6420/13/$ - see front matterPublished by Elsevier Inc. http://dx.doi.org/10.1016/j.ophtha.2013.01.018Main Outcome Measures: Demonstration of acornea.bubble popped to measure the popping pressure. Tissue obtained from these experiments was subjected tohistologic examination.

novel pre-Descemets layer (Duas layer) in the humanA Novel Pre-Descemets Lay

Harminder S. Dua, MD, PhD,1 Lana A. Faraj, MD, MJames Lowe, MD, FRCPath2

Purpose: To dene and characterize a novel pre-DDesign: Clinical and experimental study.Participants: We included 31 human donor sclerocMethods: Air was injected into the stroma of dono

in the clinical deep anterior lamellar keratoplasty procexperiments were performed: (1) creation of BB follopeeling off of the DM followed by creation of the BB, a(Duas Layer)

1 Dalia G. Said, MD, FRCS,1 Trevor Gray, MSc,2

cemets layer in the human cornea.

neal discs, including 6 controls (mean age, 77.7 years).hole globes (n 4) and sclerocorneal discs (n 21) asre with the big bubble (BB) technique. The followingd by peeling of the Descemets membrane (DM); (2)(3) creation of the BB and continued ination until the

Methods

2.5%

Ele

scanning electron microscope (JEOL, Herts, UK) and transmission

Ophthalmology Volume -, Number -, Month 2013Tissue samples were prepared for SEM and transmission electronmicroscopy after standard procedures (Appendix 1, available onlineat http://aaojournal.org). We carried out SEM using a JSM 840

2buffered glutaraldehyde.

ctron MicroscopyThe following experiments were performed on human eye bankdonor eyes:

1. creation of the BB followed by peeling off of the DM;2. peeling off of the DM followed by creation of the BB; and3. reation of the BB and continued ination till the bubble

popped to measure the popping pressure.

Tissue samples obtained from the above experiments were sub-jected to histologic examination by light, transmission and scanningelectron microscopy (SEM) and immunohistochemistry.

Air Injection

Air was injected through the epithelial surface into the stroma ofdonor whole globes (n 4) and sclerocorneal discs (n 8) andfrom the endothelial surface of sclerocorneal discs (n 13). A30-gauge needle was inserted from the limbus into the mid-peripheral stroma. For injection from the epithelial surface, donorsclerocorneal discs were mounted on a Barron articial anteriorchamber (Katena, Denville, NJ). In total, we used 25 eyes (4 wholeglobes [2 donors] and 21 sclerocorneal discs [17 donors]). Thecauses of death were infections (n 5), cardiac related (n 4),cancer (n 3), and others (n 7). Before use, all sclerocornealdiscs were maintained in organ culture in Eagles minimumessential medium with 2% fetal bovine serum for 4 to 8 weeks afterdeath. Whole globes were obtained within 48 hours of death and,after injection of air, each sclerocorneal disc was excised and thecornea examined from the posterior surface for presence of a BB.All bubbles were measured with surgical calipers.

DM Peeling

With the BB facing up, the edge of the DM close to the scle-rocorneal junction was scratched with a crescent blade or tip ofa forceps. Thus lifted, the edge of the DM was grasped with a pairof forceps and peeled centrally up and across the BB. Occasionally,the DM tore and came off in small strips. It was possible to pick upthe edge from another peripheral site to remove all of the DM.

Measuring Popping Pressure

A standard, calibrated pressure gauge (range, 0e2.5 bar) wasconnected to the side arm of a 3-way cannula. The needle forinjection was attached to 1 end of the cannula and the syringe to theother. The piston of the syringe was steadily pushed to increase thepressure until the BB popped. The pressure exerted on the air in thesyringe during injection was read directly off the gauge. After thebubble popped, the layer of tissue was grasped with a forceps andan attempt was made to mechanically peel this layer off theperiphery of the cornea in 4 samples.

Sample Preparation

Cornea with BB. For immunostaining, the BB was deated byaspirating the air with a needle inserted through the stroma into thebubble and the space relled with optimal cutting temperaturecompound before snap-freezing in liquid nitrogen. For electronmicroscopy, the entire cornea with the bubble was immersed inelectron microscopy with a JEOL 1010 microscope (JEOL).

Immunostaining

Frozen sections of tissues (Leica cryomicrotome; Wetzlar,Germany) obtained from the above experiments were stained withuorescent dyes for collagens I, IV, V, and VI and the proteo-glycans lumican, mimecan, and decorin. The adhesion moleculeCD34 and 40,6-diamidino-2-phenylindole (a uorescent stain fornuclear DNA) were used to detect cell nuclei and keratocytes,respectively (Appendix 1).

Controls

Human donor cornea samples (n 3) maintained in Eaglesminimum essential medium without air injection were used forcomparing the number of stromal layers in a given width of corneacompared with the DL.

In another 3 sclerocorneal discs, after obtaining a type-1 BB,the posterior wall of the type-1 BB was excised along its circum-ference and further injection of air carried out to determine whetheranother BB could be created.

Results

The average age of donors was 77.7 years (range, 53e94 years;median, 82 years). There were 10 females and 9 males. A BB wasobtained in all 4 whole globes and in 18 of the 21 discs. Three typesof BB were obtained. Type 1 is a well-circumscribed, centraldomeeshaped elevation measuring 7 to 8.5mm in diameter (Fig 1A,B, video clip 1), which always started in the center of the cornea andenlarged circumferentially toward the periphery (n 14). Type 2 isa thin-walled, large BB of maximum 10.5 mm diameter (Fig 1C, D,video clip 2), which, in contrast, always started as 1 or 2 smallbubbles at the periphery, enlarging centrally (coalescing) to forma large BB (n 5). The mixed type is a primary BB as in type 1,but with 1 smaller secondary bubbles as in type 2 (n 3; Fig1E, F).

With the type-1 BB it was possible to peel off the DMcompletely without deating the BB every time (Fig 2A, B videoclip 3), indicating that the posterior wall of the BB was made of theDM and an additional layer of tissue. With type-2 BB, peeling theDM resulted in deation of the bubble as soon as the edge of theBB was reached, indicating that all the air was beneath DM withno additional layer (video clip 4). With the mixed type BB,when the edge of the peeled DM reached the edge of thesecondary bubble(s), the secondary bubble(s) deated and theDM was found to be continuous with the wall of the secondarybubble(s). With continued peeling, the DM could be removedfrom the surface of the primary BB without deating it, as seenin type-1 BB.

In 5 sclerocorneal discs, a type-2 BB was rst obtained, whichcollapsed when the DM was peeled off, but in the same scle-rocorneal discs, upon further injection of air, a type-1 BB could becreated, indicating that the presence of the DM was not essential tothe formation of a type-1 BB (video clip 4). In the 4 instances wheresustained air pressure was applied to push the type-1 BB to itsmaximum dimension, the bubble extended to a 9-mm diameterand popped with a sound. For 2 of these bubbles, the poppingpressure was measured at 1.4 and 1.5 bar. For a type-2 BB, thepopping pressure was 0.6 bar. The DL could not be mechanicallypeeled off beyond a 9-mm diameter, but tension on the DL inducedne wrinkles or striae that could be seen extending across the edgeof the type-1 BB toward the periphery of the cornea (Fig 3A, B,video clip 5). Such striae could also be seen in type-1 BBs

Dua et al Novel Layer in the Human Corneaobtained after peeling off the DM, when the bubble was forcefullyinated (in the experiments designed to pop the bubble).

After obtaining a type-1 BB, when the posterior wall of thetype-1 BB was excised along its circumference (n 3 controls)and further injection of air carried out, in no case was another BBcreated, indicating that the type-1 BB is not a random separation ofa few layers of the deep corneal stroma (video clip 5).

Figure 1. All images show the sclerocorneal disc viewed from the endothelial ssclerocorneal disc and are well circumscribed and circular. A, A black circle issclera. C, D, Type-2 BBs. The bubbles extend almost to the periphery of the corstroma. Vision blue dye was applied to the endothelium. E, Mixed-type BB. Aoclock) of the sclerocorneal disc. The peripheral margin of this horse shoeeshapof the horse shoe bubble is elevated above the posterior wall of the central type-BB where the central type-1 BB is associated with a crescent-shaped, narrowerHistopathology

Light and electron microscopy revealed that the posterior wall of thetype-1 and mixed type (primary) BBs was consistently made of DMand DL (Fig 4A, B); type-2 BB was made of DM only (Fig 4C), andthose created after removing theDMweremade of DL only (Fig 4D).When the DM was partially peeled off, the difference between DL

ide. A, B, Type-1 big bubbles (BBs). The bubble occupies the center of thedrawn to delineate the periphery of the cornea, beyond which is a frill ofnea, represented by the black circle drawn in C. The needle is visible in thehorse shoeeshaped type-2 BB is seen along 10 clock-hours (from 8 to 6

ed bubble extends very close to the margin of the cornea. The posterior wall1 BB, which is seen in the hollow of the horse shoe. F, Another mixed-typetype-2 BB inferiorly. The tip of a swab points to the type-2 BB.

3

with and without DM was evident (Fig 4A). The DL thicknessmeasured a mean standard deviation of 10.153.6 m (range,

circumference (Figs 4A and 5D). Centrally, ends of broken strandscould be seen on the DL and the stromal bed (Figs 4F and 5D).

Figure 2. A, The Descemets membrane (DM) is being peeled off the surface of a type-1 big bubble (BB). The fold of the DM is clearly visible, exposing theDuas layer underneath. The bubble has not deated. B, A narrow strip of the DM is being peeled off another type-1 BB, which has not deated.

Ophthalmology Volume -, Number -, Month 20136.3e15.83 m) and the DM measured a mean standard deviation10.972.36 m (range, 7.8e13.98 m). The DL was made of 5 to 8thin lamellae of tightly packed collagen bundles running inlongitudinal, transverse, and oblique directions (Fig 4D). Incomparison, the corresponding width of the corneal stroma anteriorto DL in uninated control eyes demonstrated only 3 to 5 lamellae.The brils measured 21.702.43 nm in DL and 24.202.68 nm inthe corneal stroma (P

Dua et al Novel Layer in the Human CorneaRecent studies of specimens obtained during unsuccessfulDALK operations and from eye bank eyes have reported thepresence of residual stroma on the DM.10-12 Another recentstudy13 of eye bank eyes has demonstrated that air injectioninto the stroma induced separation of the DM with noresidual stroma in 14 of 16 corneas attempted. Our studyhas clearly demonstrated that a BB can be formed by the

Figure 4. A, Light photomicrograph of resin section stained with toluidine blue shhas been peeled off centrally to reveal the Duas layer (DL). At the periphery of thcollagen bundles (S) are seen bridging the stromal bed and posterior surface o(objective40). BeF, Transmission electron micrographs. B, The posterior wallbe made of multiple thin lamellae. Part of an endothelial cell (EC) is visible ona type-2 BB showing that it is made up of DM and attached EC only. The bandeexactly the same with the DM peeled off a type-1 BB. No split in the BZ and NBshowing multiple lamellae with collagen bundles running in longitudinal, transvmicrons. E, The DL of a type-1 BB from which the DM has not been peeled ofcollagen (arrow). Bar 1 micron. This is also evident in (D), where it appea type-1 BB showing a coiled end of a broken strand on its anterior surface (betwseparation of the DM only or by a cleavage occurring ina plane along the last row of keratocytes separating a thinbut tough layer of corneal collagen along with the DM. Thatthe DM could be completely peeled off the type-1 BBwithout deating it and that a type-1 BB could be formedafter rst peeling off the DM indicates that the DL isa distinct layer that is impervious to air. The presence of the

owing a type-1 big bubble (BB) fromwhich theDescemets membrane (DM)e cleavage, the DM can be seen on the posterior surface of the DL. Strands off the DL. A keratocyte (arrow) is seen in the anterior surface of the DLof a type-1 BB showing the DL closely applied to the DM. The DL is seen tothe posterior surface of the DM. Bar 10 microns. C, The posterior wall ofd (BZ) and nonbanded (NBZ) zones of the DM are clearly visible. This wasZ was seen in either instance. Bar 5 microns. D, The DL from a type-1 BBerse, and oblique directions. The lamellae are compactly arranged. Bar 5f. The banded zone of the DM is seen adjacent to the DL with long-spacedars as dark material in the posterior lamellae of the DL. F, The DL fromeen the 2 arrows). Some keratocyte cellular debris is also present.

5

Ophthalmology Volume -, Number -, Month 2013DM is not essential for its formation. This contrasts witha type-2 BB, which deates immediately as the DM ispeeled off from the edge of the cornea to the edge of thebubble. The corneal stroma allows air to be forced inbetween the lamellae, where numerous tiny bubblesaccumulate, rendering it opaque. When air reaches theimpervious DL, it affects a cleavage. Even at high pressure,air did not pass through this layer, which held until it burst.Type-2 BBs, on the other hand, were always formed fromthe periphery toward the center. This suggests that the DLends before the end of the DM. Air escaping posteriorlybeyond the edge of the DL at the periphery of the corneathus gains access to the plane between the DM and posteriorsurface of the DL, producing a type-2 BB. Although all ourexperiments were carried out with air injection, it isanticipated that the same types of bubbles are observedwhen viscoelastic is used instead because it would be forcedalong the same planes. The relatively tightly packedlamellae and greater space between brils in the DL

6possibly accommodating a greater amount of proteoglycans,could contribute to making it airtight.

In 2002, Hirano et al14 identied a layer attached to thedeep stroma removed by mechanical dissection. Theyattributed this to a split between banded and nonbandedlayers of the DM. With their BB technique, Anwar andTeichmann5 demonstrated that a white, semiopaque,circular ring in the cornea indicated the formation of theBB. Later, they described another less common type ofbubble with the clear edge, which they suggested wasbecause of air between the 2 layers of the DM.15

Jafarinasab et al10 in 2010 described residual stromaadherent to the DM in samples removed after attemptedBB-DALK converted to penetrating keratoplasty. Simi-larly, McKee et al11,12 created a BB in human sclerocornealdiscs by injecting air from the anterior and posterior surfaces.They also demonstrated residual stroma attached to theDM in all but 2 cases. Both studies concluded that the BBtechnique does not bare the DM in most cases, with the split

Figure 5. Scanning electron micrographs. A, The anterior surface (star) ofthe Duas layer (DL) showing parallel and regularly arranged collagenbundles. The posterior surface of the DL (triangle) shows a smooth pleatedpattern. The edge of the Descemets membrane (DM) is also seen. B, Thestromal bed of a type-1 big bubble showing a crisscross pattern with gapsrelated to passage of air (arrows). This contrasts with the appearance of theanterior surface of the DL seen in (A). C, The anterior surface of the DM.This was the smoothest of all surfaces examined. D, Strands of tissue areseen bridging the stromal bed and the anterior surface of the DL. Ends ofbroken strands are visible as small round dots (arrow). E, The end of thecleavage created by the air bubble between the DL and posterior stroma isvisible. The DL is seen as a distinct and compact layer.

he i(40,6d w

Dua et al Novel Layer in the Human Corneaoccurring within the stroma; however, in another study of 14eye bank eyes, air injection into the periphery produced DMseparation with no residual stroma.13 The latter investigatorshad inserted the needle in the peripheral 1 to 2 mm of thecornea, which is very likely to have been peripheral to thetermination of the DL, forcing air directly above theperipheral part of the DM, separating it from the DLwithout affecting a cleavage of the DL itself.

These studies complement data presented in this study,which supports our hypothesis that the BB cleaves offa distinct layer at the posterior surface of the corneal stroma,which is not residual stroma. Although the cleavageextends only 9 mm of the central cornea and is stronglyadherent thereafter, the ne wrinkles in the DL extendedbeyond the edge of the BB, suggesting that the DL extendsfurther to the periphery but not as far as the DM. When air isforced into corneal stroma the affected area becomes white.The white ring seen at the edge of the type-1 BB can beexplained by air being forced into the stroma at the cleavageplane, rendering it white. This also explains why the type-1BB is always circular and does not extend to the cornealperiphery during DALK surgery. Most DALK procedures

Figure 7. Immunostaining for collagen and proteoglycans. A, Collagen 1. Ta thin line, is the same (Alexa Fluor 555). The nuclei appear as blue dotsat 4). B, Collagen VI. The staining intensity is greater in the DL compareare no nuclei visible in the DL in either image.carried out by the BB technique are between 7 and 8.5 mmin diameter and partial thickness trephination is carried outbefore injecting air. The surgeon continues to inject air untilthe white ring reaches the trephine mark. On the basis of ourstudy, we contend that larger diameter DALKs should notbe attempted by the BB technique as the cleavage betweenDL and posterior stroma is unlikely to extend beyond 8.5mm. Instances of sudden bursting of the BB during DALKhave been experienced by us and other surgeons (unpub-lished data, Ramesh K, Glasgow, Scotland, June 2009, andDua HS and Said DG, Nottingham, England, February2011). Given the toughness (popping pressure) of the DL, itis very likely that BB bursting is a risk in type-2 BB, wherethe DM is not supported by the DL. It is also the clinicalimpression of several surgeons that eyes with DALK havestronger wounds than eyes with penetrating keratoplasty.This too can be attributed to the strength of intact DL leftbehind in the former.With lamellar keratoplasty, the occurrence of interfacehaze remains an issue. Clinical experience has taught us thatDALK and ultrathin Descemets stripping endothelial ker-atoplalsty are associated with less interface problemscompared with anterior lamellar keratopasty and Descemetsstripping endothelial keratoplasty.16-18 Hence, retention ofthe DL together with the DM and endothelium in DALK isunlikely to be associated with increased risk of interfacehaze compared with retention of the DM alone. Knowledgeof the existence of this layer and its characteristics willinuence corneal surgery; for example, the plane betweenthe DL and stroma can be exploited in generating tissue forendothelial transplant, allowing easier handling and inser-tion of the tissue because it does not tend to scroll as muchas the DM, with the DL splinting the DM. It will also helpour understanding of posterior corneal pathology such asacute hydrops in keratoconus and pre-Descemets dystro-phies. The shape and biomechanical properties of the corneaare attributed to the compact anterior lamellae and Bowmanzone. We suggest that the tough posterior DL may alsocontribute in this regard.

Because all experiments were carried out in adult eyes

ntensity of staining of the corneal stroma and the Duas layer (DL), seen as-diamidino-2-phenylindole is a uorescent stain for nuclear DNA [DAPI]ith the posterior corneal stroma (Alexa Fluor 555 and DAPI at 4). Therewith a mean age of 77.7 years and a median of 82 years,these data cannot be directly extrapolated to younger eyes.However, because the majority of DALK procedures arecarried out for keratoconus, wherein the clinical observa-tions supporting the existence of DL are evident, it isreasonable to suggest that the layer is well-dened in thatage group too. Further studies are needed to dene itscharacteristics in very young children. We were unable toexplore this question as part of this study because of thepaucity of childrens eyes that are available for research.

References

1. Muller LJ, Pels E, Vrensen GF. The specic architecture of theanterior stroma accounts for maintenance of corneal curvature.Br J Ophthalmol 2001;85:43743.

7

2. Bettelheim FA, Plessy B. The hydration of proteoglycans ofbovine cornea. Biochem Biophys Acta 1975;383:20314.

3. Bron AJ. The architecture of the corneal stroma. Br J Oph-thalmol 2001;85:37981.

4. Melles GR, Lander F, Rietveld FJ, et al. A new surgicaltechnique for deep stromal, anterior lamellar keratoplasty. Br JOphthalmol 1999;83:32733.

5. Anwar M, Teichmann KD. Big-bubble technique to bareDescemets membrane in anterior lamellar keratoplasty.J Cataract Refract Surg 2002;28:398403.

6. Manche EE, Holland GN, Maloney RK. Deep lamellar kera-toplasty using viscoelastic dissection. Arch Ophthalmol1999;117:15615.

7. Amayem AF, Anwar M. Fluid lamellar keratoplasty in kera-toconus. Ophthalmology 2000;107:769.

8. Anwar M. Technique in lamellar keratoplasty. Trans Oph-thalmol Soc U K 1974;94:16371.

9. Archila EA. Deep lamellar keratoplasty dissection of hosttissue with intrastromal air injection. Cornea 1984e1985;3:217e8.

10. Jafarinasab MR, Rahmati-Kamel M, Kanavi MR, Feizi S.Dissection plane in deep anterior lamellar keratoplasty usingthe big-bubble technique. Cornea 2010;29:38891.

11. McKee HD, Irion LC, Carley FM, et al. Residual cornealstroma in big-bubble deep anterior lamellar keratoplasty:

a histological study in eye-bank corneas. Br J Ophthalmol2011;95:14635.

12. McKee HD, Irion LC, Carley FM, et al. Donor preparationusing pneumatic dissection in endothelial keratoplasty: DMEKor DSEK? Cornea 2012;31:798800.

13. Yoeruek E, Bayyoud T, Hofmann J, et al. Comparison ofpneumatic dissection and forceps dissection in Descemetmembrane endothelial keratoplasty: histological and ultra-structural ndings. Cornea 2012;31:9205.

14. Hirano K, Sugita J, Kobayashi M. Separation of cornealstroma and Descemets membrane during deep lamellar kera-toplasty. Cornea 2002;21:1969.

15. Anwar M. Big-bubble technique. In: Fontana L, Tassinari G,eds. Atlas of Lamellar Keratoplasty. San Giovan: Fabiano;2007:12536.

16. Dapena I, Ham L, Melles GR. Endothelial keratoplasty:DSEK/DSAEK or DMEKdthe thinner the better? Curr OpinOphthalmol 2009;20:299307.

17. Bhatt UK, Fares U, Rahman I, et al. Outcomes of deep anteriorlamellar keratoplasty following successful and failed bigbubble. Br J Ophthalmol 2012;96:5649.

18. Reinhart WJ, Musch DC, Jacobs DS, et al. Deep anteriorlamellar keratoplasty as an alternative to penetrating kerato-plasty: a report by the American Academy of Ophthalmology.Ophthalmology 2011;118:20918.

Ophthalmology Volume -, Number -, Month 20138Presented at: the Societa Italiana Cellule Staminali e Supercie Oculare, VICONGRESSO S.I.C.S.S.O. Lecce, June 14e16, 2007; The Royal College

Nottingham. NG7 2UH. England, UK. E-mail: [email protected] and Financial Disclosures

Originally received: August 21, 2012.Final revision: December 7, 2012.Accepted: January 10, 2013.Available online: ---. Manuscript no. 2012-1293.1 Larry A Donoso Laboratory for Eye Research, Division of Ophthal-mology and Visual Sciences, School of Clinical Sciences, Division ofHistopathology, University of Nottingham, United Kingdom.2 Larry A Donoso Laboratory for Eye Research, Division of Ophthalmologyand Visual Sciences, School of Clinical Sciences, School of MolecularMedical Sciences, University of Nottingham, United Kingdom.of Ophthalmologists, Annual Congress 2007, Symposium on EvolvingTechniques in Corneal Surgery e Layer by Layer and EuCornea annualcongress, Milan, 2012.

Financial Disclosures:The authors have no commercial or proprietary interest in any materialsdiscussed in this article.

Supported by the Elizabeth C. King Trust.

Correspondence:Harminder S. Dua, Division of Ophthalmology and visual sciences,B Floor, Eye ENT Center, Queens Medical Center, Derby Road,

Human Corneal Anatomy RedefinedMethodsAir InjectionDM PeelingMeasuring Popping PressureSample PreparationCornea with BB

Electron MicroscopyImmunostainingControls

ResultsHistopathologyImmunohistology

DiscussionReferences