- Top of page
- Materials and method
Spontaneous chronic corneal epithelial defects have been well described in the dog.[1-3] In affected patients, these corneal defects are characterized by varying degrees of blepharospasm, corneal ulcers with nonadherent epithelial margins, and prolonged wound healing. These corneal defects in dogs share numerous clinical features with chronic or recurrent corneal erosions in people. In both canine and human patients, a plethora of treatments are available, including epithelial debridement, needle stromal puncture or grid, and superficial keratectomy, meeting with variable success rates.[4, 5]
Diamond burr debridement (DBD) has been widely advocated in the treatment of recurrent erosions in people.[6-8] Simple epithelial debridement alone may be associated with higher recurrence rates than other therapies. Stromal puncture with a needle, in either a punctate or grid pattern, may be associated with scar formation, and carries with it a risk of corneal puncture. DBD requires only inexpensive equipment, produces no scar in the cornea, and may be safer than needle stromal puncture. Like epithelial debridement and stromal puncture, DBD is easily repeated in patients that have not healed after a single procedure. Based on results of a study of 36 dogs with nonhealing ulcers, DBD with placement of a bandage contact lens (BCL) may be a safe and effective treatment of canine superficial chronic corneal erosions and may be associated with a faster healing time than grid keratotomy. An advantage of DBD over superficial keratectomy in canine patients is the ability to perform DBD without general anesthesia. Other advantages in dogs are similar to those reported in people. Histological evaluation of DBD effects in dogs showed no effect on the corneal stroma, only on the epithelium, with basement membrane involvement dependent upon duration of DBD.
Nonhealing corneal ulcers have been described in horses. In a series of 23 patients that were treated with debridement, grid keratotomy, or superficial keratectomy, time to healing averaged 20 days. Histologic evaluation of four keratectomy specimens from that study revealed a PAS-positive zone in the stroma without an epithelial basement membrane, consistent with the superficial acellular hyalinized stromal zone in dogs with superficial chronic corneal erosions. In a more recent series of 10 patients treated with grid keratotomy, average healing time is reported as 8.4 days, although this average only includes seven patients that remained hospitalized until their ulcers were fluorescein-negative, so the exact date of complete epithelialization was known. However, this average includes a healing time of 9 days for one patient that had not healed after 25 days following an initial grid keratotomy and received a second grid keratotomy, with a healing time of 9 days following the second grid. It may have been a more accurate assessment of the healing time to use 34 days as the end-point for this case, which would result in an average healing time of 12 days following the first grid keratotomy. Additionally, two cases that did not remain hospitalized until their ulcers no longer took up fluorescein stain were not included in the average healing time, because the exact healing date was not known. These two patients were noted to have been healed a recheck examination at 45 days and 35 days, respectively. Inclusion of these two patients would have resulted in an average healing time of 18 days for this study.
Nonhealing corneal ulcers can be a source of chronic pain in affected horses, and can be frustrating and expensive for clients. They are a frequent cause for referral to veterinary ophthalmologists. The safety and efficacy of DBD in horses with nonhealing corneal ulcers has not been reported. The objectives of this study were (i) to review the signalment, clinical characteristics, and outcome of horses with nonhealing corneal ulcers treated with DBD, with or without placement of a BCL; and (ii) to evaluate the role of ulcer duration, size and location on healing.
- Top of page
- Materials and method
Diamond burr debridement, which has shown promise as a therapy for chronic and recurrent corneal ulcers in humans and dogs, was evaluated as a therapy for horses with nonhealing corneal ulcers. Horses with evidence of corneal infection were not considered candidates for DBD. The average patient was a middle-aged gelding with an axial corneal ulcer with raised epithelial margins that had been present for about 1 month. In the majority of patients, no underlying etiology for the corneal ulcer had been identified. Diamond burr debridement was associated with a healing time of about 2 weeks.
Breed of affected horses was representative of the respective DACVO patient populations. As with canine patients, chronic corneal ulceration in horses appears to be a disease of middle-aged to aged patients, with an average age of 14.68 years. There was an increased incidence of chronic corneal ulcers in male horses in this study, inconsistent with the slightly increased incidence of recurrent erosions in female humans. No sex predisposition has been reported in dogs. The cause of the increased incidence of chronic corneal ulcers in male horses in this study is unknown, although it may simply represent a hospital sampling bias.
Healing time following DBD was not correlated with ulcer location, which was most commonly axial, with duration, or with size. Placement of a BCL was the only factor significantly related to healing time, with ulcers treated with a BCL taking significantly longer to heal. It is not clear that the presence of the BCL was causative in this relationship. At least two other factors confounded this result: BCLs were more likely to be placed in ulcers having a nonhealing ulcer for a longer duration than ulcers without BCLs, and BCLs were more likely to be placed following DBD performed at a University teaching hospital (by MLU). It is possible that longer duration prior to seeking treatment by a DACVO was associated with longer duration to return for recheck examination, artifacts reflecting bias of the clientele or availability of appointments in the university hospital, rather than a true delay in time to healing.
In the present study, three horses had previously been diagnosed with pituitary pars intermedia dysfunction (PPID). Pituitary pars intermedia dysfunction, commonly known as Equine Cushing's Syndrome, is a disease of aged horses that is associated with decreased corneal sensitivity. Pituitary pars intermedia dysfunction, caused by a pituitary adenoma, is the most common endocrine disorder in horses, and is characterized clinically by hirsutism. Decreased corneal sensitivity may be a mechanism underlying recurrent and chronic corneal ulceration in horses. One 20-year-old Quarter Horse gelding with PPID was enucleated after failure to heal 14 days following DBD, associated with economic constraints, as well as with humane concerns about chronic pain and poor wound healing in an aged horse who would likely tolerate loss of an eye retired to pasture. Two other horses with PPID, both aged Thoroughbred geldings, had positive bacterial cultures, although healing duration was increased relative to the mean for only one of them. Horses with PPID may represent a group at increased risk for chronic corneal ulcers, and more extensive diagnostics (i.e., corneal cytology and culture) as well as more aggressive supportive care following DBD (e.g., the addition of an antifungal) may be warranted for these patients.
Chronic ulcers included in this study fit well with the classic clinical description of “nonhealing ulcer”. The axial location recorded most frequently among ulcers in this study is also the location reported to be the most sensitive,[15, 16] although healing time was not correlated with ulcer location. In some patients, nonhealing corneal ulcers were associated with another ocular condition, including esinophilic keratitis, calcific band keratopathy, habronemiasis, and endothelial decompensation associated with glaucoma. Large differences in healing time were not noted relative to the mean healing time for all patients, although the six patients with esinophilic keratitis did have the longest healing time (19.0 days). Esinophilic keratitis has been associated with extended healing time, as long as 3–4 months in one recent study, suggesting that horses with esinophilic keratitis without secondary corneal infection may be candidates for DBD.
Complications of DBD, including infection and globe perforation, were rare. Four of 60 horses had evidence of corneal infection, three based on corneal cytology or culture collected prior to DBD and all of whom healed, and one who developed fungal keratitis post-DBD and for whom DBD was considered a failure. Only one ulcer treated with DBD, which had developed secondary to mechanical trauma from sutures placed to repair a lid laceration, progressed to perforation. Corneal ulcers associated with mechanical trauma may thus be poor candidates for DBD. In two patients for whom DBD was considered a failure, perceived delayed healing time was the basis for electing an alternate procedure (i.e., enucleation at 21 days post-DBD in a 20-year-old Quarter Horse gelding with PPID, and amnion graft at 10 days post-DBD in an 8-year-old Thoroughbred gelding), and thus actual healing time could not be evaluated. As already discussed for patients with PPID, diagnostic tests such as corneal cytology and culture may also be warranted for horses that are at risk for prolonged healing or complications, such as horses with esinophilic keratitis or horses with corneal ulcers secondary to mechanical trauma, and for horses for whom loss of an eye would be a significant impairment, such as one-eyed horses or horses required to have two eyes to compete, to rule out infection prior to DBD.
Based on results of this retrospective study, diamond burr debridement, with or without BCL placement, may be an effective treatment for nonhealing corneal ulcers in horses. The likelihood of recurrence following DBD is unknown, and would prove a valuable topic for a subsequent investigation.