Case Report: Corneal graft as surgical management for deep corneal ulceration in a horse



An 8-year-old pony was diagnosed with a progressive infected deep corneal ulcer affecting the left eye. Despite intensive medical treatment the corneal ulcer deteriorated and surgical intervention was necessary to preserve the globe. A frozen corneal graft was placed under general anaesthesia to repair the corneal defect. The corneal graft healed without complication and the patient regained good vision despite the location and size of the defect.


An 8-year-old pony was referred to the Comparative Ophthalmology Unit at the Animal Health Trust (AHT) with a five day history of severe blepharospasm and corneal ulceration affecting the left eye. Treatment prior to referral included chloramphenicol ointment q 12h. Progression of the ulcer prompted referral to the Animal Health Trust (AHT).

The pony presented with marked blepharospasm and mucopurulent discharge from the left eye. Further ophthalmic examination was performed under intravenous sedation and an auriculopalpebral block (Bupivacaine 0.25% AstraZeneca, Luton UK) on the left side.

Examination of the right eye was unremarkable. Examination of the left eye revealed marked conjunctival and deep episcleral/scleral hyperaemia. There was diffuse corneal oedema and a deep stromal ulcer with a malacic appearance medial to the central corneal axis. The ulcer bed and adjacent cornea showed marked cellular infiltration which gave the affected area a yellowish appearance (Fig. 1). Secondary reflex uveitis was present with a slightly more constricted pupil compared to the right eye. Aqueous flare and some fibrin were present in the anterior chamber of the left eye. The rest of the ophthalmic examination was unremarkable.

Figure 1.

Left eye and close up of the corneal changes. Note the corneal neovascularisation on the medial aspect and the diffuse corneal oedema. The ulcer shows malacic changes with dense cellular infiltrate. In the picture on the right hand side, detail of the deep ulcer can be seen.

The intraocular pressures were estimated with the rebound tonometer (TonoVet®) and were 25 and 12 mmHg in the right and left eye respectively.

After topical anaesthetic (proxymethacaine 0.5% Bausch & Lomb minims, Chauvin Pharmaceuticals Ltd, Surrey, UK), corneal cytology was performed. The sample was collected with a cytobrush. This revealed marked numbers of neutrophils with some intra- and extracellular coccoid bacteria and some normal keratocytes (Fig. 2). No fungal hyphae or rods were seen on the cytology slide. A diagnosis of septic ulcerative keratitis with secondary uveitis was made.

Figure 2.

Pictures of the corneal cytology. On the left hand side, a chain of cocci within eosinophilic material can be seen. On the right hand side an epithelial cell can be seen together with strands of degenerated chromatin and nuclei. Strands of cocci are also visible.

Under the same sedation a subpalpebral lavage was placed through the upper eyelid to facilitate intensive medication and decrease manipulation of the eye. Constant delivery pump infusion was attached to the lavage delivering 0.08 ml/h of chloramphenicol eye drops (Martindale Pharmaceuticals, Essex UK). Atropine eye drops were given to effect (Bausch & Lomb Minims, Chauvin Pharmaceuticals Ltd, Surrey, UK). Systemic medications included q 24h flunixin meglumine 1.1 mg/kg (Equinixin, Equine Healthcare, Norbrook® Laboratories Ltd, Newry) and q 12h trimethoprim sulfasalazine 20 mg/kg (Norodine Granules®, Equine Healthcare, Norbrook® Laboratories Ltd, Newry, UK).

During the first 24 hours of hospitalisation mild improvement in the extent of corneal malacia, and progressive corneal neovascularisation was evident. Two days after the initial consultation slight deepening of the dorsal margin of the ulcer bed was detected and this progressed to very deep ulceration (with only a small amount of stroma over Descemet's membrane) by five days. Considering the progression of the disease and risk of perforation despite intensive topical medication, surgical intervention was recommended.

A corneal graft (transplant) was recommended in order to achieve the best structural support and a relative transparent visual axis. An equine corneal specimen from the AHT Corneal Bank (CoBa) was used (Fig. 3).

Figure 3.

Intraoperative photographs. Corneal ulcer after debridement. Right lower corner: picture of globe with donor cornea waiting for complete tissue thaw.

The patient underwent general anaesthesia combined with neuromuscular blockade to ensure a centrally positioned globe. The surgery was performed with the help of an operating microscope. The bed of the corneal ulcer and walls were debrided and the corneal button was sutured to the area with simple interrupted suture pattern using nylon 9/0 (Ethicon® Division of Johnson & Johnson Medical Limited, Livingston, Scotland) (Fig. 4).

Figure 4.

Intraoperative photograph. The corneal button is being sutured with nylon 9/0 in a simple interrupted pattern. The picture on the right is the immediate postoperative picture before placing the soft contact lens.

A soft contact lens was used in order to improve ocular comfort postoperatively and to increase the contact time of the topically applied drugs.

The patient progressed well postoperatively with the corneal graft remaining in situ and the secondary uveitis controlled. The patient was discharged three days postoperatively with chloramphenicol eye drops q 6h through the lavage and flunixin meglumine 0.5 mg/kg q 24h. Re-examination after two weeks revealed the patient to be comfortable and the corneal graft to be vascularised. At this stage topical antibiotherapy was reduced to q 8h, oral anti-inflammatories were maintained and the soft contact lens was removed.

Five weeks postoperatively the patient was sedated in order to remove the corneal sutures. The 9/0 nylon sutures were removed with the help of magnifying loupes and a 25 G needle. The neovascularisation was regressing at this stage. Due to the suture removal and inevitable epithelial and stromal disruption, continuation of the chloramphenicol drops q 8h was recommended for another 10 days. At that stage the patient was comfortable and visual with no fluorescein uptake present at the site or the corneal graft, the subpalpebral lavage was removed at this stage. Mild scarring was present in the area where the surgery was performed. Three months postoperatively the owner reported the patient to be comfortable and visual.


Corneal ulceration is a common condition in horses (Gilger, 2011) and a methodical ophthalmic examination is paramount to investigate for potential aetiologies and guide the best treatment options for an individual.

The auriculopalpebral block performed on a sedated or non-sedated patient helps to perform a full ophthalmic examination with better patient compliance, especially when the patient is painful and blepharospastic.

As part of a full ophthalmic examination an estimation of the intraocular pressure can be helpful in many cases. This is not only important in cases where high intraocular pressure is suspected but also to identify hypotony which can be due to intraocular inflammation. In this case fibrin and flare was seen during the ophthalmic examination. All these clinical signs, together with the low intraocular pressure, are consistent with (or reflex) uveitis. It is generally accepted that an IOP difference of 5–8 mmHg between eyes is abnormal; this rule can be helpful when assessing ophthalmic cases.

Corneal cytology helped to rule out other causative organisms such as fungi, which should be on the list of differential diagnosis for corneal ulceration in horses in the UK (Sansom, 2005).

Medical treatment is normally preferred for corneal ulceration as this. However in cases where the depth of the ulcer progresses with the attendant risk of globe perforation surgical treatment is needed. Despite intensive medical treatment, progression of the ulcer to the deep stroma led to the decision to intervene surgically. Medical treatment in this case was eased by the use of the subpalpebral lavage system. Although we routinely place subpalpebral lavage systems through the medial lower eyelid which tends to be less mobile and provides the added benefit of increased corneal protection from the nictitans membrane, the upper eyelid was chosen due to the medial location of the corneal pathology.

Several surgical procedures could have been performed in this patient, such as conjunctival graft (with or without the use of biosynthetic material, such as porcine small intestinal submucosa, Biosis®), corneoconjunctival transposition or amnion membrane grafting (also available at the AHT) (Bussieres et al. 2004, Wilkie & Whittaker 1997, Lassaline et al. 2005).

The various pros and cons of these procedures need to be weighed up for individual patients. Conjunctival grafting will bring blood vessels to the damaged area but may not give enough support to corneal tissue especially in perforated globes. Corneoconjunctival transposition (CTT) will bring healthy transparent cornea into the defect, but this technique causes damage to otherwise healthy cornea and may lead to significant astigmatism (Ofri, 2007). Amnion corneal grafting has also been described in cases of malacic ulcerative keratitis and despite the amnion having anti-inflammatory, anti-angiogenic, antifungal and antifibrotic properties (Plummer, 2009), tissue from the AHT's CoBa to try and achieve transparency of the grafted area without damaging other areas of healthy cornea.

Frozen corneal tissue coming from the CoBa at the AHT was used as a tectonic support (i.e. scaffold) to the damaged cornea. This ‘foreign’ tissue initially becomes vascularised by the recipient. Subsequently the tissue is reorganized by the patient's own corneal cells and the vascularisation regresses leaving a semi-transparent cornea.

The use of donor tissue is increasing in veterinary medicine and the authors think that readily available stored tissue can make a difference to the way we treat patients. Corneal opacity and scarring is a major concern in horses which need to perform and the aim of the ophthalmologist presented with severe corneal disease is to try to minimize the subsequent corneal ‘haze’ as much as possible. The use of stored frozen cornea in this case demonstrates the clinical utility of this grafting procedure in a horse with severe advanced corneal disease. Although there is not much literature available on the life span of frozen corneal tissues and protocols for storage in veterinary species the experience at the AHT using CoBa tissues in dogs, cats and horses has been highly encouraging.


In order to test your understanding of this article, answer these multiple choice questions, or if you are a subscriber, go online at, and find many more multiple choice questions to test your understanding.

  • 1Which periocular block will prevent the upper eyelid movement::a. retrobulbar blockb. auriculopalpebral blockc. supraorbital blockd. none, the upper eyelid can only be controlled under general anaesthesia
  • 2Which of these clinical signs are indicative of reflex uveitis:a. miosisb. blepharitisc. high intraocular pressured. ‘a’ and ‘b’ are correct
  • 3Which of these statements is correct:a. conjunctival grafts have particular utility in a perforated globesb. corneoconjunctival transposition procedures will avoid damage of non-pathologic corneac. fungal hyphae cannot be detected by cytologyd. conjunctival grafting is a technique used to bring vascularisation to the area of damaged cornea

Answers to the above questions appear on page 47 of the print version, and as supporting information in the online version of this article at:

Supporting Information

Additional Supporting Information may be found in the online version of this article:

S1: Continuing Professional Development – Answers.