Deep lamellar endothelial keratoplasty in 10 horses


  • C. E. Plummer,

    1. Departments of Small and Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610-0126, USA
    Search for more papers by this author
  • M. E. Kallberg,

    1. Departments of Small and Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610-0126, USA
    Search for more papers by this author
  • F. J. Ollivier,

    1. Departments of Small and Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610-0126, USA
    Search for more papers by this author
  • K. P. Barrie,

    1. Departments of Small and Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610-0126, USA
    Search for more papers by this author
  • D. E. Brooks

    1. Departments of Small and Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610-0126, USA
    Search for more papers by this author

Address communications to:
C. E. Plummer
Tel.: +1 352 392-2235
Fax: +1 352 392-6125


Objective  To describe and evaluate a surgical technique utilized for the therapy of deep corneal stromal abscesses (DSA) in horses. The DSA is excised and replaced with a partial thickness corneal lamellar allograft.

Methods  A retrospective clinical study describing the indications for the surgical technique utilized and the outcomes of this procedure in 10 eyes of 10 horses.

Results  Each affected eye had a discrete DSA within the posterior stroma. An initial partial thickness semicircular corneal incision was made at the limbus, followed by anterior stromal lamellar dissection over the lesion. After excision of the DSA and replacement with a larger diameter split-thickness donor button, the anterior stroma was replaced into its original position and the initial corneal incision was repaired. All of the animals that underwent deep lamellar endothelial keratoplasty (DLEK) procedure healed appropriately and with subjectively less postoperative scarring and complications than previously described surgical approaches to DSA.

Conclusions  This procedure is an effective technique for surgical removal of DSA in horses and, in most cases, results in a visual and cosmetically acceptable globe. The advantages of this technique compared to other surgical approaches to DSA are the peripheral location of the incision, shortened anesthesia times, the resultant minimal scarring and shorter healing times associated with DLEK.


Corneal stromal abscesses (DSA) in horses are a significant disease process that has the potential to render a previously normal eye extremely painful and blind. This disease process consists of two major components, the corneal lesion itself, which consists of an accumulation of inflammatory cells and debris with or without and infectious agent or foreign body, and the secondary uveitis that follows the corneal insult.1–3 Therefore, the goals of therapy in this situation are twofold. The primary lesion must be addressed with antimicrobial medications and aggressive anti-inflammatory therapy must be instituted for the uveitis.1–3 It is the uveitis that results in damage to the interior structure of the eye thereby threatening the visual function of the globe. If medical therapy alone is insufficient to control the corneal lesion and its associated uveitis, a surgical approach to remove the offending primary lesion is indicated.

Surgical removal of a corneal abscess consists of removal of the portion of cornea wherein the lesion lies and replacement of that piece of cornea in order to maintain structural integrity of the eye. The type of surgical procedure chosen depends on the depth and location of the offending lesion. In many cases, superficial lesions may be excised and a conjunctival graft placed over the resultant defect with successful outcomes. Deep lesions often require replacement of the excised lesion with a stronger tissue. Donor cornea, that is, either fresh or frozen is utilized to repair the surgical defect in most cases. The surgical approach may be either full-thickness or partial thickness, depending upon the size and location of the abscess.4–10 If the abscess occupies the majority of the thickness of the cornea, the eye is best served by removal of the entire thickness of cornea at the diseased site via a penetrating keratoplasty (PK).4 In order to speed vascular incorporation of the donor graft and decrease the incidence of dehiscence and secondary infections at the surgical site, many surgeons opt to cover the donor PK graft with an autologous conjunctival flap.1,4,6–11 This results in a greater scar and limited vision through the surgical site, however. In an attempt to minimize local scarring and to avoid some of the potential complications of PK, partial thickness excisional procedures have been developed. Posterior lamellar keratoplasty (PLK) is a technique in which a flap of clear anterior corneal stroma is elevated over the deep abscess allowing the deep lesion occupying only posterior cornea or penetrating the anterior chamber to be excised and replaced with a partial thickness piece of donor cornea.5 The anterior stroma and epithelium is then replaced into its normal anatomic positioned and sutured in place. Since the anterior flap is in relatively close proximity to the underlying graft, a stable seal achieved by suturing in the deep layer is necessary to ensure the integrity of the repair and to prevent leakage of aqueous humor out of the eye through the two layers of cornea. The partial thickness approach to the DSA results in a strong, two-layer closure when performed properly and much less scarring than PK. There is, of course, demonstrable fibrosis at the graft site as the donor cornea is incorporated into the host tissue and along the incision sites in the anterior host cornea. Deep lamellar endothelial keratoplasty (DLEK) is a similar technique to PLK in that it is partial thickness excision and replacement.6 It was developed as a technique to limit incisions and suturing that result in corneal fibrosis in the cornea by initiating the anterior corneal incision in the periphery at the limbus.6–9 It is particularly useful for lesions in the paraxial or peripheral cornea and it shortens surgical and anesthesia times.6,7,9 The purpose of this paper is to describe in detail the DLEK surgical technique and to report the results of the procedure as performed in 10 clinical cases.


Ten horses of which there were five geldings, four mares and one stallion presented for evaluation of profound ocular pain with corneal opacities. Eight of the patients presenting were Quarter Horses, with one each of the breeds Thoroughbred and Irish Draft Horse. Patients ranged in age from 6 months to 18 years. Five affected eyes were OS and five were OD. All patients were subsequently diagnosed with deep stromal abscesses and secondary uveitis in the affected eye. An inciting incident was unknown in most of the cases, however, three were known to have had a superficial ulcer at the site prior to the development of the DSA.

The abscesses were all located in the posterior stroma and many adjacent to Descemet's membrane, if not already penetrated through that structure and into the anterior chamber (Fig. 1). The location and diameter (4–7 mm) of the DSA varied (Table 1). The decision to pursue surgical removal of the lesion was based upon the depth of the lesion alone (deep within the posterior stroma or into the anterior chamber with unaffected overlying anterior stroma), the persistence of secondary uveitis despite appropriate anti-inflammatory and mydriatic/cycloplegic medical therapy or the worsening of the secondary uveitis in the face of appropriate medical therapy. The DLEK procedure was chosen over other surgical options in most cases due to the peripheral location of the lesion.

Figure 1.

This illustration shows a posterior stromal lesion in the peripheral cornea.

Table 1.  Patient profiles, including surgical details, complications and outcomes of 10 horses undergoing DLEK for treatment of DSA
CaseSexAge (years)BreedEyeLesion locationSize of lesion (mm)Inciting incident?Sx decision
1G11QHODTemporal paraxial5Superficial ulcer (Gram + cocci, septate fungal hyphae)Depth of lesion
2M2.5QHOSDorsolateral paraxial5UnknownDepth of lesion
3G14QHODAxial6UnknownDepth, lack of vessels
4G 6QHODVentral perilimbal5Trauma, superficial ulcerDepth, worsening uveitis
5S 8QHOSVentrolateral perilimbal7UnknownDepth, worsening uveitis
6M11TBOSDorsomedial paraxial5UnknownDepth, persistent uveitis
7M6 monthsIrish DraftODTemporal perilimbal6Trauma, ulcerDepth, worsening uveitis
8G11QHOSNasal paraxial4UnknownDepth, persistent uveitis
9G18QHOSVentromedial perilimbal4UnknownDepth of lesion
10M 7QHODVentrolateral perilimbal5UnknownDepth, persistent uveitis

Initial medical therapy for all of the eyes included placement of a subpalpebral lavage (Mila, Inc, Erlanger, KY), through which natamycin (Natacyn®, Alcon Laboratories, Fort Worth, TX) (q4 h), triple antibiotic (neomycin-polymixin B sulfate-gramicidin ophthalmic solution, Bausch & Lomb® Pharmaceuticals, Tampa, FL) (q6 h), and atropine (Atrosulf-1®, 1% ophthalmic solution, Wharton, Inc., Cambridge, MA) (q6 h) were applied topically to the eye. Systemic anti-inflammatory therapy was also instituted with flunixin meglumine (Banamine, 50 mg/mL, Schering Plough, Kenilworth, NJ) 1 mg/kg PO (q12 h) and omeprazole (Gastrogard® Paste 2.28 g, Merial, Duluth, GA) 1 mg/kg (q24 h). The day prior to surgery systemic antibiotics were begun (Trimethoprim/Sulfamethoxazole 960 mg, 30 mg/kg PO BID, Mutual Pharmaceutical Co., Inc, Philadelphia, PA). Two horses were treated prior to and following surgery with systemic itraconazole (Wickliffe Pharmacy, Lexington, KY, 3 mg/kg PO (q12 h)) and one received lufenuron (Program®, Novartis Animal Health US, Greensboro, NC, 5 mg/kg PO q24 h).

The day of surgery, each animal was placed under general anesthesia and the affected eye was cleaned and prepped routinely. Neuromuscular paralysis was induced with atracurium (Tracrium®, 10 mg/mL, Glaxo Wellcome, Middlesex, UK) to facilitate eye positioning and exposure. The surgical procedure in all cases was performed with the aid of an operating microscope. Once the proper exposure was achieved and the cornea was rinsed and moistened with normal saline, an approximately two-thirds depth limbal incision was made with either a #64 or a #69 Beaver blade (BD Ophthalmics Systems, Waltham, MA). The length of the incision varied depending upon the location of the lesion and how much of the anterior cornea needed to be elevated to expose the abscess (Fig. 2). A stromal pocket was formed over the DSA with a Martinez corneal dissector with care being taken to free the periphery of the anterior flap before introducing the dissector atop the abscess (Fig. 3). Bleeding from vascularized cornea was frequently noted and was controlled with application of topical phenyleprine (2.5%, AK-Dilate™, Akorn, Inc., Buffalo Grove, IL), pressure, and/or electrocautery. The superficial corneal flap was gently retracted and the size of the abscess was measured with Jameson calipers. The borders of the abscess were delineated with a trephine (biopsy punch, size dependent on area of tissue to be removed). Attempts were made to remove the entire abscess if possible along with at least a 1-mm border of clear cornea (Fig. 4). The trephine was rotated gently over the removal site so that a groove was made, but not all the way through the cornea. A #65 Beaver blade was used to puncture the cornea within the groove and thereby enter the anterior chamber. If necessary, viscoelastic material (hyaluronate sodium, 10 mg/mL, Hylartin V™, Pfizer Animal Health, New York, NY) was injected into the anterior chamber through the perforation site to keep the eye inflated thereby protecting the interior structures of the eye from the surgical manipulations of the cornea. The viscoelastic moved the iris posteriorly, and in many cases broke down adhesions between the abscess and any retrocorneal membranes that were present, particularly in cases that had penetrated Descemet's membrane and were into the anterior chamber and adherent to the iris. Right and left corneal scissors were used to excise the delineated corneal tissue. The diseased keratectomy button was processed for aerobic and fungal culture and sensitivity, cytology, and histopathology.

Figure 2.

The initial incision is made in the perilimbal cornea and a lamellar dissection performed above the lesion to be excised. The dashed line indicates the site of the initial incision and the arrow indicates the direction of the lamellar dissection.

Figure 3.

The elevated anterior cornea flap is illustrated here.

Figure 4.

The corneal abscess is removed from the posterior cornea from underneath the anterior flap.

The anterior chamber was reformed with viscoelastic material once the stromal lesion was removed to protect the interior of the eye while the replacement steps were being performed. In some cases, dilute epinephrine was injected into the anterior chamber as well to either stem intraocular hemorrhage (n = 3) or facilitate pupillary dilation.

The donor corneal button was prepared while the diseased tissue was being excised so as to decrease the amount of time the eye was open and exposed. The replacement tissue utilized was a frozen corneal allograft. Corneas had been previously harvested from horses euthanized for non-ocular and non-neurologic reasons, such as lameness. The anterior half to two-thirds of the donor cornea including the epithelium was removed by hand dissection and discarded. Only the posterior cornea, including Descemet's membrane, was utilized. A corneal trephine 1 mm larger in diameter than the one used to excise the lesion in the host was used to obtain a circular graft from the split-thickness donor tissue. Descemet's membrane looks like cellophane and will curl outward when cut. This structure should be oriented toward the interior of the eye when it is placed within the surgical defect (Fig. 5). The large anterior corneal flap was then sutured in place so that the initial incision was repaired with 8-0 Vicryl® (violet braided polygalactin 910 suture, Ethicon, Inc., Cornelia, GA) in a simple interrupted pattern (Fig. 6). Before the incision was s completely closed, the donor graft was inserted into the excision site with Utrata forceps maneuvered gently between a space in the corneolimbal sutures. The remainder of the limbal incision was then sutured closed. The graft occasionally needed to be positioned in place by a needle inserted between the flap sutures, or by limbal injection of viscoelastic into the anterior chamber. In three cases, the surgeon opted to place the graft into the defect and suture it into place, as with a PLK, before replacing the anterior stromal flap. Four to eight simple interrupted sutures of 8-0 Vicryl® placed split-thickness were utilized to this end. Tacking the graft into place minimized the incidence of graft migration from the excision site and resulted in subjectively less corneal edema along the dissection plane, although it does require more surgical and anesthesia time for the patient. The decision of whether or not to suture the graft into the defect was based on surgeon preference.

Figure 5.

A piece of donor cornea is used to fill the defect that results from the excision of the abscess. It is illustrated here in pink. If the surgeon chooses to suture the graft into the void, this will be performed at this point before the anterior flap is replaced atop.

Figure 6.

The anterior cornea is replaced and the initial perilimbal incision is sutured closed.

Three cases had a conjunctival advancement flap placed over the limbal incision in the peripheral cornea at the conclusion of surgery in an attempt to protect the suture line and hasten vascular ingrowth. These advancement flaps covered only the perilimbal corneal incision and did not extend across the entire surgical site. In every case, the corneal graft at the excision site was visible beneath healthy anterior cornea alone. The decision of whether or not to place an advancement flap was one of surgeon preference and was made because of a subjective assessment of the poor or compromised condition of the anterior corneal flap at the level of the incision Because there is more scarring associated with the placement of an advancement flap, this additional protective measure was not employed or deemed necessary in every case. The viscoelastic was not removed from the anterior chamber. A partial temporary tarsorrhaphy was placed in all eyes to protect the surgical site during recovery.

Following surgery, topical therapy was continued as before surgery with the addition of autologous serum and in some cases, but not all, cyclosporinee ointment 0.2% (Optimmune, BID, Schering Plough, Kenilworth, NJ) to mediate graft rejection or hypertonic saline 5% (Muro 128 o/s, Bausch & Lomb, Tampa, FL) to minimize corneal edema and decrease the formation of epithelial bullae. Autologous serum was used in all cases for the first few days postoperatively, even those that received a conjunctival advancement flap, to decrease proteolytic activity induced by surgical manipulation of the cornea and to facilitate healing. One horse received topical cyclosporine and five horses received topical hypertonic saline. The differences in therapy were based upon the subjective assessment of the different clinicians attending to the individual cases. Most cases did not receive cyclosporine because of concern as to the efficacy of penetration of the drug to the intrastromal site of disease. Systemic therapy was continued as before. One horse received acetazolamide (Bedford Laboratories™, Bedford, OH) orally to aid treatment of a postanesthetic hyperkalemic periodic paralysis crisis.


Following the recovery period, all 10 horses had successful outcomes with comfortable, visual eyes (Figs. 7–9). All surgical sites healed with complete resolution and either excellent (n = 7) or good cosmesis (n = 3), and minimal scarring. Cases with excellent cosmetic results were those that had minimal to no scarring along the plane of lamellar dissection and modest fibrosis of the graft site such that the lay observer would not notice evidence of previous ocular disease without careful scrutiny. Good cosmesis was defined as mild scarring of the dissection plane and either denser fibrosis at the graft site or a larger graft scar due to the larger size of the keratectomy specimen.

Figure 7.

Case 4 at presentation. Note the superficial corneal ulcer over the focal abscess. The ulcer preceded formation of the abscess in this case.

Figure 8.

Case 4 two weeks after surgery. Blood vessels have invaded the donor graft during its incorporation. The sutures holding the graft in place are visible. The pupil is dilated and the uveitis that accompanied the corneal abscess has resolved nicely.

Figure 9.

Case 4 six weeks after surgery. The blood vessels that invaded the cornea in response to the disease and the healing process are less significantly less perfused. The opacities that remain are fibrotic and inactive. Medical therapy was discontinued after this examination.

All cases in this series developed some degree of corneal edema along the plane of dissection within the cornea, but it resolved without incident in all cases. The degree of corneal edema was more significant in cases wherein the donor graft was not sutured in place. In these 10 cases, there appeared to be no appreciable long-term difference in the cosmetic outcome between cases wherein the graft was slid into place or those into which it was sutured. Evaluation of larger numbers of each modification of the procedure may reveal a difference in the future. The main difference in the cosmetic appearance of the globes lay in the size of the lesion and the amount of anterior stroma that needed to be elevated to access the DSA. The larger the abscess, the more significant the scar and the farther into the axial cornea the DSA was located the larger the area of dissection and therefore the more fibrosis along that plane (Figs. 10–12).

Figure 10.

Case 2 at presentation. Note the focal cellular opacity in the dorsolateral cornea and the severe edema in the peripheral cornea and surrounding the lesion. There is significant uveitis present here as indicated by the circumlimbal corneal vascularization, corneal edema, and miosis.

Figure 11.

Case 2 one week after surgical excision of the lesion. The graft is clear, but swollen. Note the corneal edema that persists along the lamellar plane of dissection.

Figure 12.

Case 2 two months after surgery. There is fibrosis along the plane of lamellar dissection and at the points of suture penetration; however, over time with continued corneal remodeling, the opacities present became much less prominent and the visual axis cleared considerably.

All cases were treated postoperatively with systemic flunixin meglumine (1 mg/kg PO q12 h) and trimethoprim/sulfamethoxazole (30 mg/kg PO q12 h), as well as topical natamycin (q4 h), triple antibiotic solution (q6 h), and atropine (q6 h). The systemic antibiotics were used prophylactically due to the necessity of entering the globe during the surgical procedure and were continued for 10 days into the postoperative period. The flunixin dose was decreased slowly over the convalescence period once ocular comfort, persistent mydriasis and resolution of aqueous flare were achieved. The frequency of administration of the topical agents was decreased as well over time. Topical atropine was decreased once mydriasis was achieved. The antimicrobial agents were continued q4–6 h for the first 4 weeks after surgery after which point they were either tapered or discontinued. The horses were hospitalized for between 5 and 14 days after surgery and had daily examinations during that period. After discharge from our facility, each animal was re-examined approximately every 7–10 days until all medication was discontinued.

Postoperative medical therapy was continued for between 4 and 10 weeks, with an average of 6.7 weeks. The two cases with the extended therapy were the individuals that suffered from postoperative complications. One horse suffered from a secondary bacterial infection at the incision site and the other developed superficial keratitis at a site distant to the primary lesion. The horse that experienced infection along the incision presented with suture abscesses that appeared as white cellular opacity within the anterior stroma of the cornea along the track of each suture. An Enterococcus sp. was responsible and the addition of topical ciprofloxacin (Ciloxan® 0.3% ophthalmic solution, Alcon Laboratories) (q4 h), to which the organism was susceptible, resulted in resolution of the infection. It is impossible to know precisely how the distant keratitis that developed in the second horse relates to the primary corneal abscess, the secondary uveitis or the surgical manipulation of the globe. We suspect, however, that it was unrelated to the primary disease and was not the result of surgical introduction of an infectious agent into unaffected areas of cornea due to its mild nature and rapid resolution. All eyes in this case series have remained visual with no recurrence for the entire length of the follow-up period, which ranged from 6 months to 5 years.

The keratectomy specimens of all of the abscesses were examined histologically following excision. All lesions demonstrated a profound keratitis, wherein the cellular components were mixed and varied from suppurative to histiocytic. Neutrophilic inflammation was present in most samples, however. Foreign materials were not identified in any of the keratectomy specimens, nor were any bacterial agents. Fungal organisms were identified in 7 of the 10 samples examined. Aerobic and fungal cultures were performed on all keratectomy specimens and yielded no growth in all instances.


The traditional theory regarding the development of a corneal abscess suggests that an ulcer or microtrauma occurs in the superficial cornea that either allows an infectious agent to infiltrate the cornea or inoculates the corneal stroma with an infectious agent.1–3,10 The superficial epithelial wound then heals and seals within the cornea a microbe, microscopic agent or foreign body that attracts leukocytes and results in a visible corneal opacity, or abscess. The majority of corneal abscesses, particularly the deeper stromal lesions, are the result of infection of the cornea with a fungal organism.1–3,10,11 Even in cases where a fungal agent was not identified on histopathologic examination of an abscess keratectomy specimen, fungal DNA is present in most instances.11 Fungal agents have an affinity for the Type VI collagen of Descemet's membrane and are often found congregating within and around it.1 This is the case in particularly aggressive cases of ulcerative fungal keratitis as well, in which the organism burrows vertically through the superficial tissue to its target.1,2 These findings suggest that DSA should be considered to be of a fungal etiology, and treated as such, until proven otherwise by the identification of another etiologic agent or mechanism, even in the absence of demonstrable fungal agents.

Abscesses within the cornea may occur at different levels of the stroma and can change position or consolidate over the course of the disease and its treatment. They may occupy the anterior or posterior aspects of the cornea and in some cases may even penetrate through Descemet's membrane into the anterior chamber. Superficial lesions are often treated effectively medically as medications have less tissue through which to penetrate to affect a result at the site of interest and, more importantly, the vascular response that invades the cornea from the limbus is most effective in the anterior portion of the cornea. In order for a cornea stromal abscess to heal medically and ultimately resolve, the lesion itself must completely vascularize. In many cases of deep stromal abscesses, or those that occupy the posterior stroma, the lesions are incompletely vascularized during healing. Blood vessels may progress over the top of the abscess rather than infiltrate it. During healing of stromal abscesses, uveitis is a frequent occurrence and internal ocular structures may be at risk of damage. Medical therapy for corneal abscesses is often prolonged and may extend from 4 weeks to 6 months or more. The treatment period may be shortened significantly with surgical removal of the lesion.1,4–8 The decision to surgically remove an abscess is also based upon its location within the cornea, the degree of medical control of the uveitis and the extent of the vascular response in the cornea to the lesion.1–9,11 If the uveitis associated with a DSA does not respond to anti-inflammatory and mydriatic/cycloplegic therapy, if it worsens despite appropriate anti-inflammatory therapy or if it recurs or worsens with the tapering or discontinuation of anti-inflammatory medication, surgical removal of the primary lesion, the DSA, is indicated.1–9,11

Keratoplasty procedures that address only the diseased portion of the cornea, leaving the healthy portion of the host cornea intact, are now the preferred method of surgical therapy for focal corneal disease.9,12–14 In the realm of human corneal surgery, PLK and DLEK were developed for patients with endothelial dysfunction in order to decrease the postoperative complication rate and improve the postoperative visual function.12–16 In the case of a stromal abscess, the endothelium is not the primarily diseased tissue. Its close proximity to DSAs in the posterior cornea and the affinity fungal agents have for Descemet's membrane, however, make it a necessary surgical target.

Removal of only the affected posterior cornea allows for a strong two-layer closure without the need for a conjunctival flap overlying the corneal graft, which results in a clearer visual axis. Procedures that limit incisions made into the clear cornea and that require fewer absorbable sutures within the visual axis will ultimately result in less corneal scarring.9 Fewer corneal sutures equates to shorter surgical and anesthesia times, which in turn are beneficial to the overall health and recovery of the horse. It may be advantageous to utilize continuous suture patterns to further limit surgical times and suture reaction.

Corneal transplantation is a viable and successful surgical technique in the horse.1–9,11 Full thickness PK may be performed for melting ulcers, descemetoceles, iris prolapses and full thickness stromal abscesses.4,8,9 DLEK and PLK are split thickness penetrating keratoplasties utilized for deep stromal abscesses with clear overlying anterior stroma.5–9 Potential complications include incision leakage, secondary infections at the surgery site, particularly abscesses along the suture lines, edema and fluid accumulation along the planes of lamellar dissection, and graft misalignment or migration in instances where the button was not tacked into place.4–8 As with any corneal transplant, 3–5 days after implantation, the graft begins to swell as part of the rejection reaction initiated by the host. The donor graft remains transparent for up to 7 days after surgery, but eventually becomes vascularized as the host incorporates the donor tissue into the structure of the eye. The resultant scar is typically vascularized initially and eventually becomes opaque. Over time with continued remodeling of the cornea, the scar will become less dense (Table 2).

Table 2.  Surgical details, histopathologic findings and results of 10 cases that underwent DLEK as treatment for corneal stromal abscess
CaseLength of time before sx (weeks)Graft sutured in place?Conjunctival flap?Histopathology findingsComplicationsResults
13NoYesSuppurative, histiocytic keratitis, no organismsNoneResolution, visual, good cosmesis, granulation tissue at ulcer site, mild posterior synechia
22NoNoSuppurative, histiocytic keratitis, fungal hyphaeSuture abscess (Enterococcus) – added topical cipro; persistent po uveitisResolution, visual, good cosmesis, slightly smaller globe
31YesYesKeratitis, fungal hyphaeNoneResolution, visual, good cosmesis, axial endothelial fibrosis
43YesYesSuppurative keratitis, fungal hyphaeNoneResolution, visual, excellent cosmesis
52NoNoKeratitis, fungal hyphaeHYPP crisis after anesthesia; none ocularResolution, visual, excellent cosmesis
63NoNoNeutrophilic keratitis, fungal hyphaeNoneResolution, visual, excellent cosmesis
72YesNoSuppurative keratitis, no organismsNoneResolution, visual, excellent cosmesis
82NoNoSuppurative keratitis, fungal hyphaeDistant superficial keratitis – added topical CsAResolution, visual, excellent cosmesis
91NoNoSuppurative, lymphoplasmacytic keratitis, no organismsNoneResolution, visual, excellent cosmesis
103NoNoNeutrophilic, lymphohistiocytic keratitis, fungal hyphaeNoneResolution, visual, excellent cosmesis

The DLEK procedure is associated with shorter surgery and treatment times than the PK.6,7,9 Surgical and anesthesia times are also shorter with the DLEK compared to the PLK, although postoperative treatment times are rather similar between these two partial thickness procedures. The shorter length of both the procedure and the convalescence period of the DLEK can have dramatic effects on the animal's welfare and may reduce morbidity and mortality associated with longer techniques and can have significant financial benefits for the owner. The procedure is not without potential complications, however, aside from those noted above. Although it was not observed in any of the cases in this series, it is theoretically possible to iatrogenically introduce a causative infectious agent into unaffected areas of the cornea with the Martinez dissector as the instrument contacts tissue above the discrete abscess that may be infected and then pushes it into new territory. This makes it possible for satellite lesions to occur adjacent to the primary lesion. Also, the larger area of corneal dissection results in larger potential area for diffuse corneal fibrosis. In most cases in this series this was insignificant, however. The scarring that occurs at the graft site alone following DLEK is less than that that occurs with focal PLK grafts; however, the potential for larger, albeit more faint, areas of fibrosis exists. In all cases to date, however, the postoperative corneal scarring after DLEK has been acceptable to both horse and owner and has not resulted in any behavioral evidence of diminished vision.

Corneal transplants in horses are associated with high success rates, good visual outcomes, and shorter treatment times than medical therapy alone. Stromal abscesses are a significant disease process in the horse and often require surgical intervention to restore an animal's vision and comfort. The DLEK procedure is an effective form of corneal transplantation for treatment of DSAs which results in a structurally sound globe and minimal corneal scarring (Fig. 13).

Figure 13.

Case 6 one year after surgery. The postoperative appearance of this eye is excellent. There is residual faint fibrosis along the initial plane of lamellar dissection. The fibrosis at the graft site is more significant; however, it did not interfere with the horse's functional vision.