Doris Wu 1* Sara M. Smith 2* Jessica M. Stine 3,4 Tammy M. Michau 3 Thomas R. Miller 4 Samantha L. Pederson 1 Kate S. Freeman 1

DOI: 10.1111/vop.12556 ORIGINAL ARTICLE Treatment of spontaneous chronic corneal epithelial defects (SCCEDs) with diamond burr debridement vs combination diamond burr debridement and superficial grid keratotomy Doris Wu 1* Sara M. Smith 2* Jessica M. Stine 3,4 Tammy M. Michau 3 Thomas R. Miller 4 Samantha L. Pederson 1 Kate S. Freeman 1 1 Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO, USA 2 University of Pennsylvania School of Veterinary Medicine, Philadelphia, PA, USA 3 BluePearl Veterinary Partners, Tampa, FL, USA 4 Tampa Bay Veterinary Specialists, Largo, FL, USA Correspondence Kate S. Freeman Email: ktsfreeman@gmail.com Abstract Objective: To evaluate the efficacy of diamond burr debridement (DBD) vs a combination of diamond burr debridement with superficial grid keratotomy (DBD+SGK) for the treatment of spontaneous chronic corneal epithelial defects (SCCEDs) in dogs. Procedure: Medical records of dogs diagnosed with SCCEDs from three different institutions that received a DBD or DBD+SGK between 2003 and 2015 were reviewed. Age, breed, sex, history of a previous SCCED, procedures performed, time to healing, and complications were statistically analyzed. Results: One hundred and ninety-four dogs met the inclusion criteria. Eighty-two of 106 eyes (77.4%) received a DBD and healed following the first treatment (13.3 4.9 days to recheck, range 2-27). Sixty-eight of 88 eyes (77.3%) received a DBD+SGK and healed following the first treatment (15.4 5.0 days to recheck, range 5-45). No significant difference in healing outcome was found between the two treatments (P = 1). For SCCEDs that healed after a single treatment (n = 150), complications occurred in 13.3% (n = 20) of eyes with no difference in complications between the DBD and DBD+SGK groups (P =.86). Thirty-five of 44 eyes (80.0%) healed after the second treatment (16 8.2 days from second treatment to third visit, range 5-47); nine of 44 eyes (20.0%) were not healed (12 6.2 days from second treatment to third visit, range 5-25). The second treatment method did not influence healing rates (P =.64). Conclusions: DBD and DBD+SGK are equally effective treatment methods for canine SCCEDs. No differences in complication rates after one treatment were observed between DBD and DBD+SGK. KEYWORDS canine, corneal ulcer, diamond burr debridement, indolent ulcer, spontaneous chronic corneal epithelial defect, superficial grid keratotomy *Co-first authors. Veterinary Ophthalmology. 2018;1 10. wileyonlinelibrary.com/journal/vop 2018 American College of Veterinary Ophthalmologists 1

2 WU ET AL. 1 INTRODUCTION Spontaneous chronic corneal epithelial defects (SCCEDs), also known as indolent ulcers or nonhealing ulcers, are noninfected, epithelial defects with a redundant, nonadherent epithelial border. They are further characterized by an intact corneal stroma and the presence of an anterior stromal hyalinized acellular membranous zone (HAZ). 1-6 SCCEDs were first described in the Boxer dogs and have since been described in other canine breeds, occurring most commonly in middle-aged to older patients. 4,5,7,8 SCCEDs fail to heal due to a lack of normal epithelial wound healing and are the most common type of refractory corneal ulcers in veterinary medicine, often present for weeks to months prior to presenting to a veterinary ophthalmologist. 1,3 A variety of medical treatments aimed at modulating the wound healing process have been reported including topical serum, oral and topical tetracycline, epidermal growth factor, substance P, aminocaproic acid, and polysulfated glycosaminoglycans. However, these topical treatments alone are often not successful in resolving canine SCCEDs. 5,8-13 Procedural and surgical treatments include corneal debridement (cotton-tipped applicator, diamond burr), grid keratotomy, anterior stromal puncture, thermal cautery, cyanoacrylate tissue adhesive, nictitating membrane flap, and superficial keratectomy. 6,14-19 For treatment of canine SCCEDs, epithelial (cotton-tipped applicator) debridement alone has reported success rates from 20%- 84%, with an overall success rate of approximately 50%, 1 but only 25% healing at first recheck of 7-10 days. 14 Bandage contact lenses have also been used as a primary treatment modality or as an adjunct to procedural or surgical therapies. 6,11,15,20 Diamond burr debridement was first reported for the treatment of corneal erosions in humans in 1983 and is one of the most commonly used methods of manual epithelial debridement. 21 A review of 40 SCCEDs that received a single treatment of DBD with placement of a bandage contact lens had a 70% healing rate at first recheck (7 days) with 92.5% healing rate by second recheck (15 5 days). 6 Another study revealed that DBD eyes were significantly more likely to have healed by three weeks when compared to SGK. 22 However, a prospective study of 27 Boxer dogs that received SGK had a 100% success rate of healing after a single treatment (7-10 days healing time), 15 while multiple punctate keratotomy had a reported success rate of 68-88%. 11,16 Stanley s 1998 study comparing SGK, superficial keratectomy, and cotton swab debridement revealed that 75% of dogs receiving a SGK healed by their first recheck visit at 10-14 days and 87% healed with only a single treatment. The high success rate of SCCED healing by SGK was suggested to be due to penetration of the superficial stroma to alter the extracellular matrix and promote epithelial attachment. 14 Histologic evaluation of the cornea immediately following DBD showed partial removal of the basement membrane with no stromal involvement. 23 A recent study using light and electron microscopy following DBD showed significant changes in the superficial stromal hyalinized acellular zone (HAZ) that may be responsible for the high healing rates associated with DBD alone. 24 Determining the most efficient and effective treatment for canine SCCEDs remains an ongoing goal in veterinary ophthalmology. With many treatment modalities currently available, the goal is to establish a treatment method or combination of methods that provide the highest success rate of SCCED healing with minimal post-treatment complications. It has been reported that treatments addressing stromal changes have higher success rates of healing SCCEDs than epithelial debridement alone. 1,6,14-16,18 Thus, the aim of this retrospective study was to compare the efficacy of diamond burr debridement (DBD) alone to the efficacy of DBD combined with superficial grid keratotomy (SGK) for the treatment of SCCEDs in dogs. The hypothesis was that combination DBD+SGK would increase the overall healing rates compared with DBD alone as it was predicted that DBD+SGK may disrupt the HAZ more effectively than DBD alone. 14,24 Secondary objectives included the evaluation of signalment, medical history, subsequent treatments, and concurrent topical therapy that may influence healing. Investigation into a possible relationship between post-treatment complications and the treatment method after the first treatment was also evaluated. 2 MATERIALS AND METHODS 2.1 Data collection Medical records of canine patients presenting to Colorado State University Veterinary Teaching Hospital (Fort Collins, CO) from 2003 to 2015 and to Blue Pearl Veterinary Partners (Tampa, FL) and Tampa Bay Veterinary Specialists (Largo, FL) from 2004 to 2010 were retrospectively reviewed. Included were cases with a diagnosis of SCCED that received a diamond burr debridement (DBD) or a combination of a diamond burr debridement with superficial grid keratotomy (DBD+SGK). Ophthalmic examinations were performed by the attending veterinary ophthalmologist or by an ophthalmology resident supervised by a veterinary ophthalmologist. The diagnosis of SCCED was based upon the following characteristic clinical signs: a superficial corneal ulcer lacking stromal involvement, a lip of nonadherent epithelium surrounding the ulcer site that expanded

WU ET AL. 3 with cotton-tipped applicator (CTA) debridement, no evidence of corneal cellular infiltrate, and a halo-fluorescein dye stain pattern. 1 Patients were excluded from the study if the eye had received a previous DBD or SGK, but were not excluded if the eye received a previous CTA debridement. Other exclusion criteria included lack of at least one documented follow-up with a veterinarian or no fluorescein stain performed at the follow-up examination. Patients diagnosed with concurrent ocular (eg, conjunctivitis, keratoconjunctivitis sicca, chronic superficial keratitis, pigmentary keratitis, corneal dystrophy or degeneration, endothelial degeneration, previous phacoemulsification, uveitis, glaucoma, distichia, ectopic cilia, trichiasis, eyelid mass, euryblepharon, local foreign bodies), or systemic disease (eg, diabetes mellitus, hypothyroidism, hyper- or hypoadrenocortism, allergic dermatitis, immune-mediated disease, systemic neoplasia) were excluded. Patients receiving topical (eg, steroid, tacrolimus, or cyclosporine) or oral immunosuppressive therapy before or after SCCED treatment were also excluded. They were excluded as topical glucocorticoids can delay corneal wound healing by affecting corneal epithelial healing rates and both topical glucocorticoids and calcineurin inhibitors reduce corneal neovascularization. 25-28 Patients with SCCEDs occurring in the same or contralateral eye at a separate visit had only the first documented SCCED included in the analysis. Additionally, in patients with bilateral SCCEDs at initial presentation, the eye that was healed first was included in the analysis. Information obtained from the medical records included the following: age, breed, sex, duration of clinical signs prior to evaluation by a veterinary ophthalmologist, prior topical therapy, prior CTA debridement, ophthalmic examination findings (Schirmer tear test, intraocular pressure), procedure performed at the specialty institution (DBD or DBD+SGK), use of a bandage contact lens, post-treatment medical therapy, time to recheck examination(s), results of follow-up examination(s), necessity for additional treatments (procedural or medical), and post-treatment complications. While tear production and intraocular pressure measurements were obtained, these data were not available or performed on all cases and thus not subjected to statistical analysis. SCCEDs are typically present longer than 7 days; however, if an eye fits all the classic clinical signs at presentation (halo-fluorescein stain pattern, loose epithelial edge, easily debrided with a CTA), the eye was not excluded with an uncertain lesion duration. Post-treatment therapy included in the statistical analysis comprised of topical atropine, topical sodium chloride (NaCl), topical Terramycin (oxytetracycline hydrochloride with polymyxin B sulfate; Pfizer Inc., New York, NY, USA), topical serum, and the use of a bandage contact lens. Terramycin is commonly used in dogs with SCCEDs for antibiotic properties and has been shown to significantly shorten the healing time for refractory corneal ulcers when used in conjunction with CTA and SGK compared to dogs not receiving Terramycin. 8 Additional medical treatment and adjunct therapies such as topical antibiotics other than Terramycin, oral nonsteroidal anti-inflammatory drugs (NSAIDs), oral tramadol, or the use of an E-collar were commonly utilized but these data were not subjected to statistical analysis. 2.2 Procedures The specifics of the procedures used in this retrospective study were at the discretion of the managing ophthalmologist at the time of treatment; however, all DBD and DBD+SGK were performed similarly. The corneal surface was anesthetized with several drops of proparacaine hydrochloride ophthalmic solution, USP 0.5% (Falcon Pharmaceuticals, Inc., Fort Worth, TX, USA), and aseptically prepared with a diluted betadine solution (1:50). The nonadherent epithelium was removed with sterile CTA debridement. DBD was then performed with a hand-held battery-operated rotating diamond burr with a 3.5-mm spherical medium grit tip (Alger Company, Inc., Lago Vista, TX, USA). The grit tip was steam sterilized following each use. The diamond burr was passed over the stromal ulcer bed in a combination of gentle circular, horizontal, and vertical motions of the burr tip, typically for the length of approximately one minute. Those patients receiving a DBD+SGK had the DBD procedure performed as described above, followed by use of a sterile 22-, 25-, or 27-gauge needle to create fine grid lines, which were approximately 0.5-1 mm apart and extended 2-3 mm beyond the normal epithelial rim. The lines penetrated approximately 5%-10% into the anterior stroma, resulting in visible linear scratching of the corneal surface (Figure 1). A bandage contact lens was placed post-procedure at the discretion of the veterinary ophthalmologist. Animals were classified as healed post-treatment if there was no evidence of fluorescein stain uptake and no evidence of corneal infection as determined by a veterinary ophthalmologist at the subsequent recheck examination. This was typically associated with a smooth epithelial surface with varying degrees of stromal opacity and often superficial corneal vascularization between the ulcer region and the nearest limbus. Eyes were considered not healed if fluorescein stain was still present at the following recheck visit. A second treatment (Treatment 2) was performed in eyes not healed after the first treatment (Treatment 1). Treatment 2 was categorized into one of five different treatment combination methods (Figure 2) taking into consideration the method performed at Treatment 1: (i) DBD+SGK at 1st with DBD+SGK at 2nd, (ii) DBD at 1st with DBD at 2nd, (iii) DBD at 1st with DBD+SGK at

4 WU ET AL. analyzed separately from eyes following the second treatment (Treatment 2) as eyes healed after Treatment 1 were not treated again. First, univariate relationships between each predictor and healing were considered. Chi-squared or Fisher s exact test was used to evaluate if treatment type (DBD or DBD+SGK), sex, breed, history of previous SCCED, or medications were associated with healing status at the subsequent recheck. Logistic regression was used to evaluate if age or time to recheck was associated with healing status at subsequent recheck. Values were considered statistically significant if the P-value was <.05. For the healing after first treatment and complications after first treatment, a multiple logistic regression model was used to test for an effect of treatment controlling for institution and additional covariates. Due to the small sample sizes, multiple logistic regression was not attempted for healing after the second treatment. FIGURE 1 Appearance of a canine spontaneous chronic corneal epithelial defect (SCCED) following a DBD+SGK treatment. Note the visible linear scratching of the corneal surface 2nd, (iv) DBD+SGK at 1st with DBD at 2nd, (v) DBD or DBD+SGK at 1st with medical management at 2nd. For statistical analysis, as only 5 eyes underwent medical therapy at the second treatment, Treatment 2 medical therapy was combined without separating DBD at 1st and DBD+SGK at 1st into two different categories. Again, animals were classified as healed or not healed based on the same criteria used for Treatment 1. Post-treatment complications were based on ophthalmic examination findings and categorized into mild, moderate, or severe. Mild complications included persistent pain at recheck (blepharospasm) requiring additional oral analgesia and subjective ophthalmic findings of corneal edema and fibrosis beyond expected healing changes. Moderate complications included diffuse corneal fibrosis, novel corneal granulation tissue, novel superficial corneal melanosis, and microbulla formation. Microbulla formation was only noted after treatment as a result of corneal healing and not from underlying pre-existing endothelial disease. Severe complications included keratomalacia, stromal lysis, and the need for enucleation. Eyes healed following the first procedure (Treatment 1) that developed post-treatment complications were evaluated separately from eyes that had not healed after Treatment 1. 2.3 Statistical analysis Statistical analysis was performed using R 3.3.1 (The R Foundation, Vienna, Austria). For all analyses, data for eyes following the first treatment (Treatment 1) were 3 RESULTS A total of 194 dogs (194 SCCEDs) were included in the study. Mean age was 8.7 years ( 2.6 years) with a range 1.5-18 years. Boxers were the most common breed (n = 57; 29.4%) followed by mixed breeds (n = 29; 14.9%), Labrador Retrievers (n = 16; 8.2%), Boston Terriers (n = 10; 5.2%), Golden Retrievers (n = 7; 3.6%), Corgis (n = 6; 3.1%), and English Bulldogs (n = 5; 2.6%). The remainder of the individual breeds each represented less than 2.0% (n 4) of the total population and were grouped together into other (n = 64; 33%). Forty-seven percent (n = 92) of dogs were spayed females and 2.1% (n = 4) were intact females. Thirty-eight percent (n = 73) were neutered males and 12.9% (n = 25) were intact males. Prior to specialist examination, the median duration of clinical signs was 21 days with a range of 1 to 180 days (n = 184). The duration of clinical signs was unknown for 10 eyes. 3.1 Results after first treatment: 194 eyes One hundred and six SCCEDs received a DBD and 88 SCCEDs received a DBD+SGK at the time of initial examination. Distribution of treatment method by specialty institution for the initial visit is listed in Table 1. At the recheck examination (second visit), 82 of 106 eyes (77.4%) that received a DBD as the first treatment were healed (13.3 4.9 days to recheck at second visit, range 2-27) and 68 of 88 eyes (77.3%) that received a DBD+SGK as the first treatment were healed (15.4 5.0 days to recheck at second visit, range 5-45). The success of healing was not significantly different for eyes treated with a DBD compared to eyes treated with the DBD+SGK technique

WU ET AL. 5 14 12 12 11 Healed Not healed 10 NUMBER OF EYES 8 6 6 5 4 3 3 2 2 1 1 0 (i) DBD+SGK 1st; DBD+SGK 2nd (ii) DBD 1st; DBD 2nd (iii) DBD 1st; DBD+SGK 2nd (iv) DBD+SGK 1st; DBD 2nd (v) DBD or DBD+SGK 1st; Medical 2nd TREATMENT METHODS 0 FIGURE 2 Healing status by treatment method for eyes after Treatment 2 (n = 44). DBD, Diamond burr debridement; DBD+SGK, Diamond burr debridement and superficial grid keratotomy. (i) DBD+SGK 1st; DBD+SGK 2nd = eyes treated with DBD+SGK for Treatment 1 and Treatment 2 (ii) DBD 1st; DBD 2nd = eyes treated with DBD for Treatment 1 and Treatment 2 (iii) DBD 1st; DBD+SGK 2nd = eyes treated with DBD for Treatment 1 then DBD+SGK for Treatment 2 (iv) DBD+SGK 1st; DBD 2nd = eyes treated with DBD+SGK for Treatment 1 then DBD for Treatment 2 (v) DBD or DBD+SGK 1st; Medical 2nd = eyes treated with DBD or DBD+SGK for Treatment 1 and then medical treatment only for Treatment 2 TABLE 1 Distribution by specialty institution of initial procedures performed for the treatment of SCCEDs (n = 194) Institution Treatment 1: Total (%) CSU (n = 84) DBD = 74 (88%) DBD+SGK = 10 (12%) TBVS (n = 93) DBD = 15 (16%) DBD+SGK = 78 (84%) FVS (n = 17) DBD = 17 (100%) DBD+SGK = 0 (0%) Percentages (%) calculated within each institution CSU, Colorado State University Veterinary Teaching Hospital; TBVS, Tampa Bay Veterinary Specialists; FVS, Florida Veterinary Specialists; DBD, diamond burr debridement; DBD+SGK, diamond burr debridement with superficial grid keratotomy. Total = total of DBD or DBD+SGK at each institution. % = percentage of total population of SCCEDs for either DBD or DBD+SGK at each institution. (P = 1) based on the Chi-squared test. There was a significant positive relationship between healing and time to second recheck based on logistic regression (P =.002) [OR 1.13; CI 95%: 1.04-1.23]. However, using multiple logistic regression with healing as the response and accounting for institution and time to the second recheck, the success of healing was not significantly different for eyes treated with a DBD compared to eyes treated with the DBD+SGK technique (P =.99). There was no significant relationship between healing and age based on logistic regression (P =.88) [OR 0.99; CI 95%: 0.87-1.13]. 3.2 Results after second treatment: 44 eyes Forty-four SCCEDs, 24 that received a DBD and 20 a DBD+SGK at the first treatment, were not healed at the second visit and subsequently received a second treatment. Evaluation of the second treatment involved categorization into one of five different treatment combination methods as outlined in the materials and methods. The categories were as follows: (i) DBD+SGK at 1st with DBD+SGK at 2nd, (ii) DBD at 1st with DBD at 2nd, (iii) DBD at 1st with DBD+SGK at 2nd, (iv) DBD+SGK at 1st with DBD at 2nd, (v) DBD or DBD+SGK at 1st with medical management at 2nd (Figure 2). Of the 20 eyes that underwent DBD+SGK for the first treatment and did not heal at the second visit, 15 eyes received a second DBD+SGK resulting in 80.0% of SCCEDs healed by the third visit (14.1 5.9 days from the second treatment to the third recheck visit, range 5-28). Of the 24 eyes treated that underwent DBD for the first treatment and did not heal at the second visit, 14 eyes received a second DBD resulting in 79.0% of SCCEDs healed by the third recheck visit (16.8 10.3 days from the second treatment to the third recheck visit, range 7-47). Of the 24 eyes that underwent a DBD for the first treatment and did not heal at the second recheck, 8 eyes received a DBD+SGK resulting in 75.0% of SCCEDs healed by the third recheck visit (11 4.7 days from the second treatment to the third recheck visit, range 5-18). Of the 20 eyes that underwent DBD+SGK for the first treatment and did not heal at the second recheck, 2 eyes received a DBD resulting in 50.0%

6 WU ET AL. of SCCEDs healed by the third recheck visit (15.0 2.8 days from the second treatment to the third recheck visit, range 13-17). Of the 5 eyes treated medically at the second treatment regardless of whether DBD or DBD+SGK was performed as the first treatment, 100% had healed by the third recheck visit (20.6 9 days from the second treatment to the third recheck visit, range 14-31). Regardless of the treatment method used for the second treatment, 35 eyes (80.0%) were healed by the third visit (16 8.2 days from the second treatment to the third visit, range 5-47) and 9 eyes (20.0%) were not healed (12 6.2 days from the second treatment to the third visit, range 5-25). There was no significant association between healing and time to third recheck based on logistic regression (P =.11) [OR 1.11; CI 95%: 0.98-1.33]. Based on logistic regression, there was also no significant association between healing and age (P =.95) [OR 1.01; CI 95% confidence 0.76-1.37]. Fisher s exact test showed no significant difference in healing rate between the 5 different treatment combination methods (P =.64). Further treatment and time to healing for the 9 eyes that had not healed after the second treatment were not further investigated. 3.3 Variables Frequency of healing following the first treatment and associated values correlated with sex, breed, and previous history of SCCED are summarized in Table 2. Medical therapies prescribed after the first treatment and frequency of healing at the second visit are summarized in Table 3. Medical therapies prescribed after the second treatment and frequency of healing at the third visit are summarized in Table 4. None of the variables analyzed were found to impact healing status after the first or second treatment based on Fisher s exact test (P >.05). 3.4 Post-treatment complications For SCCEDs that healed after a single treatment (n = 150; DBD = 82, DBD+SGK = 68), complications occurred in 13.3% (n = 20) of eyes. Using a Fisher s exact test, there was a significant difference (P =.003) in complications between eyes treated with DBD (21%, n = 17) and those treated with DBD+SGK (4.4%, n = 3). However, after using multiple logistic regression which accounts for institution, there was no difference in complications between the DBD and DBD+SGK treatment groups after the first treatment (P =.86). As only one eye experienced severe complications (DBD+SGK treatment group) after initial healing, no further statistical calculations were performed to investigate whether severe complications were associated with either treatment techniques. Table 5 summarizes post-treatment complications by severity after the first treatment. No microbial culture was performed on any of these eyes. TABLE 2 Evaluation of variables potentially influencing SCCED healing status following the first treatment for all cases (n = 194) Variable evaluated Healed (n) Not healed (n) n (% of total) P-value Sex FI 2 2 4 (2.1).44 FS 72 20 92 (47.4) MI 21 4 25 (12.9) MC 55 18 73 (37.6) Breed Boston Terrier 7 3 10 (5.2).57 Boxer 43 14 57 (29.4) Corgi 3 3 6 (3.1) English Bulldog 3 2 5 (2.6) Golden Retriever 6 1 7 (3.6) Labrador Retriever 14 2 16 (8.2) Mix 24 5 29 (14.9) Other 50 14 64 (33) History of SCCED Yes 3 1 4 (2.1).94 No 58 16 74 (38.1) Unknown 89 27 116 (59.8) FI, female intact; FS, female spayed; MI, male intact; MC, male castrated; n, total of healed and not healed eyes for each variable evaluated. % of total = percentage of total spontaneous chronic corneal epithelial defects (SCCEDs) population (n =194). A P-value of <.05 would indicate significant association with SCCED non-healing following the first treatment. The P values are based on using Fisher s exact test.

WU ET AL. 7 TABLE 3 Evaluation of influence of adjunctive therapies prescribed following first treatment (either DBD or DBD+SGK) on SCCED treatment outcome (n = 194) Treatment Healed (n) Not healed (n) n (% of Total) P-value BCL Yes 38 13 51 (26.3).57 No 112 31 143 (73.7) Atropine Yes 59 12 71 (36.6).16 No 91 32 123 (63.4) NaCl Yes 14 1 15 (7.7).2 No 136 43 179 (92.3) Terramycin Yes 10 3 13 (6.7) 1 No 140 41 181 (93.3) Serum Yes 9 4 13 (6.7).5 No 141 40 181 (93.3) BCL, bandage contact lens; NaCl, topical sodium chloride; Terramycin, oxytetracycline hydrochloride with polymyxin B sulfate; n = total of healed and not healed eyes for each treatment evaluated. % of total = percentage of total SCCED population (n =194). A P-value of <.05 would indicate a significant association with SCCED nonhealing following the first treatment. The P values are based on using Fisher s exact test. The topical treatments used following the first treatment did not significantly affect healing of SCCEDs at the follow up second recheck visit. TABLE 4 Adjunctive therapies prescribed at the second visit if the SCCED was not healed after the first treatment and their associated significance for healing at the third recheck examination (n = 44) Treatment Healed (n) Not healed (n) n (% Total) P-value BCL Yes 10 3 13 (29.5) 1 No 25 6 31 (70) Atropine Yes 9 3 12 (27.3).69 No 26 6 32 (72.7) NaCl Yes 1 0 1 (2.3).2 No 35 8 43 (97.7) Terramycin Yes 2 1 3 (6.8).5 No 33 8 41(93.2) Serum Yes 4 0 4 (9.1).57 No 31 9 40 (90.9) BCL, Bandage contact lens; NaCl, topical sodium chloride; Terramycin, oxytetracycline hydrochloride with polymyxin B sulfate; n = total of healed and not healed eyes for each treatment evaluated; % of total = percentage of total SCCED population undergoing second treatment (n =44). Eyes were treated with either a DBD, DBD+SGK or continued on medical management only at the second treatment. A P-value of <.05 would indicate significant association with SCCED nonhealing by the third recheck examination. The adjunct treatments used following the second treatment did not significantly affect healing of SCCEDs at the follow up third recheck visit. 4 DISCUSSION This study compared the efficacy of two treatment methods, DBD and DBD+SGK for the treatment of SCCEDs in dogs, the variables that may affect healing outcome, and reported post-procedure complications. The results of this study show that the success of SCCED healing after one procedure is not significantly different for eyes treated with DBD compared to eyes treated with the combination DBD+SGK. Even when using healing as the end point and accounting for institution and time to the second recheck, the success of healing was not significantly different between eyes treated with DBD and DBD+SGK. A single treatment of DBD+SGK in this study (77.3% healed, mean 15.4 5.0 days) resulted in similar healing outcomes as utilizing grid keratotomy alone in Stanley s 1998 study (75% healed, mean 13.4 5.1 days), 14 but was lower than the 100% healing rate of SGK reported by Wooff et al in Boxer dogs after a single treatment. 15 Differences in healing rates may be due to the diverse patient population, case selection, differences in protocols between institutes, and the retrospective nature of this study. Forty-four of 194 SCCEDs that received either a DBD (n = 24) or a DBD+SGK (n = 20) at the first treatment, were not healed at the follow-up second examination and subsequently received a second treatment. The second treatment was categorized into one of five treatment combinations, and regardless of the treatment technique performed at the second treatment, 35 eyes (80.0%) were healed and 9 eyes (20.0%) were not healed at the follow-up third visit. Statistical analysis showed that the success of SCCED healing was not significantly different between the 5 different treatment combinations. These results suggest that outcome of healing is not significantly different if either DBD or DBD+SGK or medical therapy is pursued in nonhealing SCCEDs after the first treatment. However, four out of five eyes for which medical therapy was pursued as the second treatment had developed complications after the first treatment. Due to the small sample size, statistical analysis was not performed. One of the four cases developed bullous TABLE 5 Post-treatment complications by severity in SCCEDs that healed after the first treatment (n = 150) Complications Treatment None Mild Moderate Severe n (% of Total) DBD 65 24 3 0 82 (55%) DBD+SGK 65 2 0 1 68 (45%) DBD, diamond burr debridement; DBD+SGK, diamond burr debridement with superficial grid keratotomy. n = total of healed eyes for each complication after the first treatment. % of total = percentage of total SCCED population that healed after the first treatment (n = 150).

8 WU ET AL. keratopathy and the remaining three developed keratomalacia; therefore, these corneal defects were no longer considered SCCEDs. These cases were treated with aggressive medical treatment instead of a repeat procedure as both DBD and SGK are contraindicated for all non-scced type corneal defects including infected ulcers. 29 The reported prognosis for recovery from a debridement procedure is considered favorable in the absence of infection or severe corneal endothelial degeneration. 30 There was a positive relationship between healing and time to second recheck with the longer the time to recheck, the higher the likelihood of healing. This finding suggests utilizing longer recheck intervals (ie, mean >14.8 5.2 days) in clinical practice, as animals are more likely to be healed by that time. Sex, breed, age, and the use of topical medications or a bandage contact lens post-treatment are all variables that may affect healing status of canine SCCEDs. Studies have shown that the addition of a bandage contact lens, collagen shield, or nictitating membrane flap can increase the success rate of epithelial debridement. 11 Based on results presented in this study (Tables 3 and 4), 51 of 194 eyes (26.3%) received a bandage contact lens after the first treatment and 13 of 44 eyes (29.5%) after the second treatment. However, this had no significant impact on healing for either treatment method after the first or second treatment. With the retrospective nature of the study, it was often unclear from the medical records whether the bandage contact lens was still present at recheck examinations or at what time the contact lens may have been removed from or fallen out of the patient s eye. None of the other variables analyzed (sex, breed, and post-treatment topical medications) were found to influence healing status after the first or second treatment. Only seven of 194 eyes (3.6%) were treated with CTA debridement prior to the first treatment (four prior to DBD and three prior to DBD+SGK); six healed after the first treatment and one after the second treatment. Although CTA debridement has been reported to successfully heal SCCEDs, 1,14 statistical analysis was not performed to evaluate whether CTA debridement influenced healing in this study due to the small number of eyes undergoing CTA debridement prior to the first treatment. As in previous studies evaluating SCCED treatments, all eyes with concurrent ocular disease were excluded from this study as concurrent disease processes and medications could affect healing. 25-28,31-35 For eyes that healed following the first treatment (n = 150), the frequency of post-treatment complications was not significantly different between eyes that were treated with a DBD compared to those receiving a DBD+SGK when accounting for institutions. This study found a total of one case (0.52% of SCCEDs, one DBD+SGK) that developed severe complications of keratomalacia and stromal lysis following healing after the first treatment. Therefore, no further statistical calculations were performed to investigate whether severe complications were associated with the initial treatment technique. It is important to remember that DBD and DBD+SGK procedures result in a necessarily large, significant corneal wound that creates an expected response of mild ocular discomfort, localized corneal edema, neovascularization, and corneal fibrosis. 14,36 Previous studies utilizing DBD and SGK have reported rare complications including varying degrees of keratomalacia, corneal perforation, permanent fibrosis, and infection. 9,17,22,37 The use of a bandage contact lens has also been associated with infectious keratitis. 38 Due to the retrospective nature of the study, the details of bandage contact lens use was often unclear, and therefore, statistical analysis was not performed to determine whether bandage contact lens use was associated with the keratomalacia cases documented in this study. Post-treatment complications that developed for eyes treated twice were not investigated or statistically analyzed. This was due to numerous confounding factors and variables such as individual susceptibility to delayed healing, contact lens placement, and time to second treatment that may have contributed to post-treatment complications but were not associated with treatment number or method. Additionally, complications following the first treatment may have influenced the need for a second treatment. Therefore, a combination of these factors may have influenced complication severity. Determination of post-treatment complication was subjective especially for those categorized as mild. These mild complications consisted of corneal edema, fibrosis, and pain beyond the expected healing changes, for which different clinicians likely have variable expectations of what are considered normal findings post-treatment and may grade such changes differently. The lack of predefined criteria to categorize posttreatment complications at the time of recheck was therefore a limitation of this study. Potentially confounding variables, various sample sizes, and a lack of predefined objective secondary end point criteria are also limitations of the retrospective nature of this study, illustrating the benefit of a future controlled prospective study. 5 CONCLUSION Diamond burr debridement (DBD) and a combination of DBD+SGK are both successful in the treatment of canine SCCEDs, with no significant differences in healing outcomes or complications between the two treatment methods. These results are in line with a recent study that reported disruption and thinning of the abnormal hyalinized acellular zone (HAZ) by DBD which is also a suggested mechanism

WU ET AL. 9 for SGK. 1,14,24 Disruption of the HAZ is likely the most critical element in the treatment of SCCEDs to allow for better epithelial attachment and successful wound healing. ACKNOWLEDGMENT The authors would like to thank Ann Hess for performing the statistical analysis. ORCID Tammy M. Michau 7851 REFERENCES http://orcid.org/0000-0001-8161-1. Bentley E. Spontaneous chronic corneal epithelial defects in dogs: a review. J Am Anim Hosp Assoc. 2005;41:158-165. 2. Bentley E, Abrams GA, Covitz D, et al. Morphology and immunohistochemistry of spontaneous chronic corneal epithelial defects (SCCED) in dogs. Invest Ophthalmol Vis Sci. 2001;42:2262-2269. 3. Moore PA. Diagnosis and management of chronic corneal epithelial defects (indolent corneal ulcerations). Clin Tech Small Anim Pract. 2003;18:168-177. 4. Gelatt K, Samuelson D. Recurrent corneal erosions and epithelial dystrophy in the boxer dog. J Am Anim Hosp Assoc. 1982;18:453-460. 5. Murphy CJ, Marfurt CF, McDermott A, et al. Spontaneous chronic corneal epithelial defects (SCCED) in dogs: clinical features, innervation, and effect of topical SP, with or without IGF- 1. Invest Ophthalmol Vis Sci. 2001;42:2252-2261. 6. Gosling AA, Labelle AL, Breaux CB. Management of spontaneous chronic corneal epithelial defects (SCCEDs) in dogs with diamond burr debridement and placement of a bandage contact lens. Vet Ophthalmol. 2013;16:83-88. 7. Kirschner S, Niyo Y, Betts D. Idiopathic persistent corneal erosions: clinical and pathological findings in 18 dogs. J Am Anim Hosp Assoc. 1982;25:84-90. 8. Chandler HL, Gemensky-Metzler AJ, Bras ID, et al. In vivo effects of adjunctive tetracycline treatment on refractory corneal ulcers in dogs. J Am Vet Med Assoc. 2010;237:378-386. 9. Kirschner S, Brazzell R, Stern M. The use of topical epidermal growth factor for the treatment of nonhealing corneal erosions in dogs. J Am Anim Hosp Assoc. 1991;27:449-452. 10. Regnier A, Cazalot G, Cantaloube B. Topical treatment of nonhealing corneal epithelial ulcers in dogs with aminocaproic acid. Vet Rec. 2005;157:510-513. 11. Morgan R, Abrams K. A comparison of six different therapies for persistent corneal erosions in dogs and cats. Vet Comp Ophthalmol. 1994;4:38-43. 12. Miller W. Use of polysulfated glycosaminoglycans in treatment of persistent corneal erosions in dogs - a pilot clinical study. Vet Med. 1996;91:916-922. 13. Eaton JS, Hollingsworth SR, Holmberg BJ, et al. Effects of topically applied heterologous serum on reepithelialization rate of superficial chronic corneal epithelial defects in dogs. J Am Vet Med Assoc. 2017;250:1014-1022. 14. Stanley RG, Hardman C, Johnson BW. Results of grid keratotomy, superficial keratectomy and debridement for the management of persistent corneal erosions in 92 dogs. Vet Ophthalmol. 1998;1:233-238. 15. Wooff PJ, Norman JC. Effect of corneal contact lens wear on healing time and comfort post LGK for treatment of SCCEDs in boxers. Vet Ophthalmol. 2015;18:364-370. 16. Champagne E, Munger R. Multiple punctate keratotomy for the treatment of recurrent epithelial erosions in dogs. J Am Anim Hosp Assoc. 1992;28:213-216. 17. Turner A, Blogg J. Multiple striate keratotomy: a treatment for corneal erosions caused by epithelial basement membrane disease. Aust Vet J. 1997;75:27-30. 18. Bentley E, Murphy CJ. Thermal cautery of the cornea for treatment of spontaneous chronic corneal epithelial defects in dogs and horses. J Am Vet Med Assoc. 2004;224:250-253. 19. Bromberg NM. Cyanoacrylate tissue adhesive for treatment of refractory corneal ulceration. Vet Ophthalmol. 2002;5:55-60. 20. Schmidt GM, Blanchard GL, Keller WF. The use of hydrophilic contact lenses in corneal diseases of the dog and cat: a preliminary report. J Small Anim Pract. 1977;18:773-777. 21. Buxton JN, Constad WH. Superficial epithelial keratectomy in the treatment of epithelial basement membrane dystrophy. Cornea. 1987;6:292-297. 22. Sila GH, Morreale RJ, Lorimer DW, et al. A retrospective evaluation of the diamond burr superficial keratectomy in the treatment of spontaneous chronic corneal epithelial defects in dogs from 2006 to 2008 (abstract). Vet Ophthalmol. 2009;12:404. 23. da Silva EG, Powell CC, Gionfriddo JR, et al. Histologic evaluation of the immediate effects of diamond burr debridement in experimental superficial corneal wounds in dogs. Vet Ophthalmol. 2011;14:285-291. 24. Dawson C, Naranjo C, Sanchez-Maldonado B, et al. Immediate effects of diamond burr debridement in patients with spontaneous chronic corneal epithelial defects, light and electron microscopic evaluation. Vet Ophthalmol. 2017;20:11-15. 25. Boneham GC, Collin HB. Steroid inhibition of limbal blood and lymphatic vascular cell growth. Curr Eye Res. 1995;14:1-10. 26. Petroutsos G, Guimaraes R, Giraud JP, et al. Corticosteroids and corneal epithelial wound healing. Br J Ophthalmol. 1982;66:705-708. 27. Lipman RM, Epstein RJ, Hendricks RL. Suppression of corneal neovascularization with cyclosporine. Arch Ophthalmol (Chicago, Ill : 1960) 1992; 110: 405-407. 28. Benelli U, Ross JR, Nardi M, et al. Corneal neovascularization induced by xenografts or chemical cautery: inhibition by cyclosporin A. Invest Ophthalmol Vis Sci. 1997;38: 274-282. 29. Maggs DJ. Cornea and Sclera. In: Maggs DJ, Miller PE, Ofri R, eds. Slatter s Fundamentals of Veterinary Ophthalmology, 4th edition. St. Louis: Elsevier Sanders; 2008:261-276,10. 30. Jones ML, Gilmour MA, Streeter RN. Use of grid keratotomy for the treatment of indolent ulcer in a llama. Can Vet J. 2007;48:416-419. 31. Guidera AC, Luchs JI, Udell IJ. Keratitis, ulceration, and perforation associated with topical nonsteroidal anti-inflammatory drugs. Ophthalmology. 2001;108:936-944.

10 WU ET AL. 32. Lin J, Rapuano C, Laibason P, et al. Corneal melting associated with use of topical nonsteroidal anti-inflammatory drugs after ocular surgery. Arch Ophthalmol. 2000;118:1129-1134. 33. Congdon NG, Schein OD, Von Kulajta P, et al. Corneal complications associated with topical ophthalmic use of nonsteroidal antiinflammatory drugs. J Cataract Refract Surg. 2001;27:622-631. 34. Chen JJ, Tseng SC. Corneal epithelial wound healing in partial limbal deficiency. Invest Ophthalmol Vis Sci. 1990;31:1301-1314. 35. Chen W, Zhao K, Li X, et al. Keratoconjunctivitis sicca modifies epithelial stem cell proliferation kinetics in conjunctiva. Cornea. 2007;26:1101-1106. 36. Cook C, Wilcock B. Clinical and histological study of canine persistent superficial corneal ulcers. In: Twenty-Sixth Annual Meeting of the American College of Veterinary Ophthalmologists. 1995;193. 37. Peiffer R, Gelatt K, Gwin R. Superficial keratectomy in the management of indolent ulcers of the boxer cornea. Canine Pract. 1976;3:31-33. 38. Dohlman CH, Boruchoff SA, Mobilia EF. Complications in use of soft contact lenses in corneal disease. Arch Ophthalmol. 1973;90:367-371. How to cite this article: Wu D, Smith SM, Stine JM, et al. Treatment of spontaneous chronic corneal epithelial defects (SCCEDs) with diamond burr debridement vs combination diamond burr debridement and superficial grid keratotomy. Vet Ophthalmol. 2018;00:1-10. https://doi.org/10.1111/ vop.12556