Nictitating membrane resection in the horse: A comparison of long-term outcomes using local vs. general anaesthesia
Reasons for performing study: Neoplasia, for which surgical excision is a frequent treatment, is the most common disease of the equine nictitating membrane. There is little long-term follow-up information available to the practitioner regarding the long-term effects of nictitating membrane excision on ocular health. No information is available to compare recurrence of primary neoplasia of the nictitating membrane after excision with local or general anaesthesia.
Objectives: To evaluate the long-term complications of nictitating membrane resection in horses; recurrence of neoplasia of the nictitating membrane when nictitating membrane resection is performed under local vs. general anaesthesia and if the method of anaesthesia used to permit resection of the affected membrane influences the recurrence of neoplasia of the nictitating membrane after complete nictitating membrane resection.
Methods: Records of 26 horses receiving resection of the nictitating membrane for primary neoplasia of the nictitating membrane 1999–2009 were reviewed. Clinical examination findings, surgical procedure, anaesthesia type, histopathological findings and details of adjunctive treatment were recorded. Owners were contacted via telephone regarding post operative outcomes. Data were analysed using a Fisher's exact test (P<0.05).
Results: The most common long-term complication of nictitating membrane excision was mild ocular discharge. Squamous cell carcinoma was the most frequent histopathological diagnosis. Recurrence of neoplasia was uncommon (2/26 horses). No significant difference in the number of horses experiencing recurrence of neoplasia was detected between groups receiving general anaesthesia vs. those receiving local anaesthesia.
Conclusions: Resection of the nictitating membrane in horses following local anaesthesia is not associated with increased risk of recurrence of neoplasia compared with excision under general anaesthesia. Resection of the nictitating membrane is not associated with any long-term ocular side effects and can be an effective modality for cure of primary neoplasia of the nictitating membrane in selected cases.
The nictitating membrane plays an important role in ocular health by acting to protect the globe and aid in both production and distribution of the precorneal tear film. The nictitating membrane is also a common site of ocular neoplasia in the horse, with squamous cell carcinoma (SCC) being the most frequent neoplasm (Blodi and Ramsey 1967; Lavach and Severin 1977). A wide variety of treatments for SCC have been evaluated, including surgical resection, cryotherapy, interstitial radiation, immunotherapy, photodynamic therapy and chemotherapy (Hilbert et al. 1977; Grier et al. 1980; Frauenfelder et al. 1982; Walker et al. 1986; King et al. 1991; McCalla et al. 1992; Theon and Pascoe 1995; Mosunic et al. 2004; Bosch and Klein 2005; Hewes and Sullins 2006; Ollivier et al. 2006; Rayner and Van Zyl 2006; Giuliano et al. 2008; Payne et al. 2009). Previous reports conflict on the ability of surgical resection of the nictitating membrane alone to effect a permanent cure for neoplasia of the nictitating membrane. A recent study from the UK suggests that surgical resection of the nictitating membrane is not associated with recurrence of neoplasia, while previous studies from the UK and USA suggest that surgery with no adjunctive therapy is associated with higher rates of recurrence (Walker et al. 1986; Dugan et al. 1991; Mosunic et al. 2004; Payne et al. 2009).
Standing sedation using local anaesthesia for surgical procedures in the horse has become more commonplace as drugs that allow both sedation and analgesia without ataxia, toxicity or cardiorespiratory compromise become more widely available. Techniques for complete excision of the nictitating membrane under local anaesthesia have been previously described (Millichamp 2005; Harper 2009; Payne et al. 2009). There are currently no studies comparing the long-term outcomes of complete nictitating membrane excision for treatment of primary neoplasia of the nictitating membrane performed under local anaesthesia vs. general anaesthesia or documenting the long-term complications of nictitating membrane excision on ocular health. The aims of this study were to evaluate the long-term complications of nictitating membrane resection in horses, to evaluate recurrence of neoplasia of the nictitating membrane when nictitating membrane resection is performed under local vs. general anaesthesia and determine if method of anaesthesia influences the recurrence of neoplasia of the nictitating membrane after nictitating membrane resection.
Materials and methods
Selection of cases
The study design was a retrospective study. Records of all horses presenting to The Ohio State University Veterinary Medical Centre between January 1 1999 and January 1 2010 were searched using the key words ‘third eyelid’ or ‘nictitating membrane’ and ‘excision’ or ‘resection’. Records were obtained through Health Information at The Ohio State University Veterinary Medical Centre. Horses were eligible for inclusion in the study if the following criteria were met: a diagnosis of suspected primary neoplasia limited to the nictitating membrane, resection was expected to be curative, complete excision of the nictitating membrane was performed by the Comparative Ophthalmology service and histopathological examination of the excised tissue was performed. Horses were excluded if the owner was unable to be contacted for follow-up information, the nictitating membrane had been excised prior to evaluation, concurrent ocular disease was noted, with the exception of SCC, at additional ocular sites or gross extension of the lesion beyond the nictitating membrane was visible.
Medical records were searched to obtain the following information: age, breed, sex, laterality, description of any ocular lesions, surgical procedure, any adjunctive treatment performed, post operative complications and histopathological findings. Owners were contacted by telephone to obtain information about time to recurrence, development of subsequent ophthalmic disease, vision deficits, development of neoplastic disease in other sites and satisfaction with the surgical procedure.
Descriptive data analysis was performed1. For continuous data, normality was assessed using the Shapiro-Wilk test. Normally distributed data was described using mean and standard deviation (s.d.) and non-normally distributed data was reported by median and interquartile range. A Fisher's exact test was used to compare the proportion of cases with recurrent neoplasia between anaesthetic groups.
A total of 38 records meeting the inclusion criteria were retrieved; however, 12 cases were subsequently excluded because the owners could not be reached to obtain follow-up information. Twenty-six cases were included in the data analysis. Surgery was performed unilaterally in 21 cases (13 left eyes and 8 right eyes) and bilaterally in 5 cases for a total of 31 nictitating membranes excised. The mean age was 13.5 years (s.d. ± 6.7 years) with 10 mares, 14 geldings and 2 stallions. Breeds included Appaloosa (3), Arabian (2), Belgian (6), Clydesdale (4), Grade (1), Morgan (1), Paint/Spotted Saddle Horse (4), Quarter Horse (2), Thoroughbred (2) and Hanoverian (1). Mass lesions of the nictitating membrane were described in all 31 eyes. Additionally, a total of 7 horses had additional ocular lesions suspected to be SCC at the time of nictitating membrane excision, including 5 limbal masses and 2 ulcerative eyelid lesions. Surgery was performed with the horses under general anaesthesia in 16 cases and under local anaesthesia with i.v. sedation in 10 cases. Clinicians tended to select local anaesthesia for small lesions that could be palpated and localised to the nictitating membrane and when there were no other ocular lesions that necessitated concurrent treatment. All horses included in the study with non-nictitating membrane lesions that were suspected to be SCC had surgery under general anaesthesia.
For cases where nictitating membrane excision was performed under local anaesthesia, routine sedation using i.v. detomidine or a combination of xylazine/torbugesic was administered. To facilitate akinesia of the eyelids, an auriculopalpebral nerve block was performed using 3–5 ml of 2% mepivicaine (Carbocaine-V, 2% mepivicaine hydrochloride USP)2. Topical 0.5% proparacaine (Proparacaine hydrochloride 0.5% ophthalmic solution USP)3 was applied to the cornea and conjunctival surfaces and 5–10 ml of 2% mepivicaine was injected into the base of the nictitating membrane (Millichamp 2005). All cases were aseptically prepared using a 1:20 to 1:50 dilution of povidone iodine solution (Vetadine 1% povidone iodine solution)4. The surgical procedure was performed similarly in all cases regardless of anaesthetic category. Two towel clamps were affixed to the leading margin of the nictitating membrane and a curved Mayo scissor was used to sharply excise the nictitating membrane at its medial and lateral margins followed by a single cut across the base, always taking care to remove the nictitating membrane cartilage and glandular tissue in its entirety. The bulbar and palpebral conjunctiva at the base of the nictitating membrane was closed in 29 of 31 surgical procedures using absorbable 5-0 or 6-0 suture in a simple continuous pattern in an attempt to prevent herniation of the retrobulbar fat. In the remaining 2 surgical procedures, the conjunctiva was not closed, including one case each performed under local and general anaesthesia.
Adjunctive therapy was used to treat the region surrounding the nictitating membrane in 5 cases at the time of surgery. All cases receiving adjunctive therapy also underwent general anaesthesia. Three horses received cryotherapy using liquid nitrogen spray at the base of the nictitating membrane, while 2 horses received beta radiation at the base, medial or lateral insertion of the nictitating membrane. Six horses received beta radiation of the lateral limbus/cornea for lesions that were suspicious for SCC at the time of surgery. Surgery was performed under general anaesthesia in all horses receiving beta radiation.
Histopathology was performed in all cases. A diagnosis of SCC was made in 24 nictitating membranes, carcinoma in situ in 2 nictitating membranes and lymphoma, basal cell carcinoma, squamous papilloma, epithelial hyperplasia, lymphoid hyperplasia in one nictitating membrane each. Histopathological examination of tissue taken from the horses with SCC revealed that 3 had incomplete excision, 2 demonstrated intravascular invasion and one had intra-lymphatic invasion. All cases with incomplete excision were performed under general anaesthesia. All 3 cases with evidence of invasion were also diagnosed as complete excision. Of the cases with incomplete excision, 2 horses had concurrent adjunctive therapy at the time of surgery (cryotherapy) and one horse received no adjunctive therapy. One horse with intra-lymphatic invasion was treated with a series of 4 injections with 1% 5-Fluorouracil (5-Fluorouracil 1% solution)5 beginning 2 weeks post operatively, while the 2 horses with intravascular invasion were not treated with any adjunctive therapy.
At the time owners were contacted for follow-up information, 17 horses were known to be alive and 9 horses were dead. The reasons for death were not related to the nictitating membrane lesion in 7 horses and included orthopaedic disease, colic, old age, cardiac disease and recurrent perineal SCC as causes of death. A single horse was subjected to euthanasia for recurrence of ocular SCC approximately 2 years post operatively and one horse died spontaneously with neurological disease suspected, but not confirmed, to be associated with a recurrence of ocular SCC approximately 5 years post operatively.
Two horses developed lesions that could be considered recurrences of their primary nictitating membrane neoplasia (SCC in both cases). One horse that was bilaterally affected and also exhibited intravascular invasion unilaterally on histopathology recurred ipsilaterally 11 months post operatively and was subsequently treated with topical 1% 5-Fluorouracil ointment. This horse was subsequently subjected to euthanasia for a recurrence in the contralateral eye some months later. A second horse developed a suspicious recurrence that was not confirmed via histopathology at 6 weeks post operatively. This horse was treated with topical 1% 5-Fluorouracil for 8 weeks which induced regression of the lesion with no reported recurrences. This same horse was diagnosed histopathologically as a completely excised SCC but did receive adjunctive cryotherapy at the nictitating membrane base as part of the original surgical procedure.
Two horses developed presumptive SCC lesions post surgery that probably represented new tumour formation given the length of time between surgical excision and development of the lesion. One horse was noted by the owner to have a mass in the region of the nictitating membrane 10 years after nictitating membrane excision. The owner declined further diagnostics or treatment due to the advanced age of the horse. A second horse was noted by the owner to have a mass in the region of the nictitating membrane 5.5 years post operatively that slowly enlarged over a one-year period until the horse died of progressive neurological disease.
The median time since surgery with no evidence of recurrence in the remaining 22 horses was 6.3 years with an interquartile range of 3.4–8.2 years. No recurrences were noted in the 10 horses in the local anaesthesia group and 2 recurrences noted in the 16 horses in the general anaesthesia group. There was no significant difference in number of horses experiencing a recurrence of nictitating membrane neoplasia between horses that received surgical excision of the nictitating membrane under local anaesthesia vs. those operated on under general anaesthesia (P = 0.508).
The only long-term complication reported by owners was a discharge classified as mild that necessitated cleaning the periocular surface once daily or less and was noted to range in character from serous to yellow/white. Of the horses without evidence of recurrence, none were reported to require veterinary care for ocular disease. Although no owners deemed additional veterinary attention necessary, follow-up examinations on all cases were not performed as part of the study, thus it is not possible to rule out small recurrent lesions or the presence of subclinical disease. All owners reported they would choose to have the procedure performed again.
Although the nictitating membrane plays an important role in ocular health, neoplastic disease may necessitate its amputation. The aims of this study were 2-fold: first, to determine the long-term effects on ocular surface health of removing the nictitating membrane and second to compare the frequency of recurrence of primary nictitating membrane neoplasia after nictitating membrane excision under local vs. general anaesthesia. The major limitation of this retrospective study is a reliance on owner perception, memory and assessment rather than examination by a veterinarian or veterinary ophthalmologist. The authors acknowledge the inherent limitations this poses in the data, but still emphasise the valuable information that can be gleaned from this study. With the exception of 2 horses that suffered recurrences of SCC and the 2 horses that developed putative novel SCC lesions, no owner in this study reported needing to seek veterinary care for an ocular problem post operatively. The results of this study suggest that clinically significant, long-term complications associated with nictitating membrane excision are minimal.
The most frequently reported complication of nictitating membrane excision in this study was a mild ocular discharge, most likely resulting from accumulation of dust, debris and mucus in the ventral conjunctival fornix. Owners should be warned of this potential complication prior to surgical excision of the nictitating membrane. Notably, no disease of the cornea, particularly ulcerative or nonulcerative keratitis, was reported by owners in this study. Excision of the canine nictitating membrane has been associated with an increased risk of developing keratoconjunctivitis sicca (KCS) with resultant loss of corneal transparency, so it is important to note that a similar association has not yet been identified in the horse (Morgan et al. 1993). Schirmer tear tests were not performed as part of this study; however, there were no clinical signs of KCS as assessed by owners. Whereas prolapse of the orbital fat has previously been reported as a complication of nictitating membrane excision, this complication was not observed in this study (Giuliano 2010). The authors advise primary closure of the conjunctival wound in order to decrease the risk of orbital fat herniation.
The inherent risks of general anaesthesia and its associated recovery make standing surgical procedures desirable in the horse (Parviainen and Trim 2000). This is the first study to compare the frequency of recurrence of nictitating membrane neoplasia after nictitating membrane excision performed in horses under general or local anaesthesia. These results suggest that surgical excision of the nictitating membrane, when performed under standing sedation and local anaesthesia, is not associated with any increased risk of recurrence of primary neoplasia, particularly SCC. Although some literature suggests that surgical excision alone is associated with higher rates of recurrence of SCC, the results of this study suggest that surgical excision in selected cases can be curative (King et al. 1991). Careful case selection followed by histopathology seems important in optimising case outcome. The most important criterion for case selection prior to performing excision of the nictitating membrane is the presence of a well demarcated lesion whose margins are readily palpable with visibly normal nictitating membrane tissue surrounding the lesion. Smaller lesions may be more amenable to surgical resection under local anaesthesia, as would lesions where no adjunctive therapy (such as beta radiation, which can be time consuming) would be performed. Resection of larger lesions where obtaining complete excision is challenging or lesions where adjunctive therapy is performed would best be performed in horses under general anaesthesia. The low frequency of recurrence in this study may be related to adherence to careful case selection as described above.
The importance of histopathology in determining the need for adjunctive post operative therapy and prognosis cannot be overemphasised. Interestingly, SCC was suspected to be the cause of the lesion in 30/31 nictitating membrane in this case series; however, only 24/31 nictitating membrane were histologically diagnosed as SCC. Histopathology provides valuable information about the need for future monitoring and adjunctive therapy and is recommended in every case where the nictitating membrane has been excised, with particular attention to the margins of the lesion.
The results of this study suggest that the long-term complications of nictitating membrane resection in the horse are minimal and limited to mild ocular discharge. Nictitating membrane neoplasia can be successfully treated with nictitating membrane resection performed under general or local anaesthesia, and performing the procedure standing in appropriately selected cases does not increase the frequency of recurrence of neoplasia. The extent of the initial lesion should influence the clinician's selection of general or local anaesthesia and may also influence the likelihood of recurrence.
Authors' declaration of interests
No conflicts of interest have been declared.
1 SPSS 17.0, SPSS Inc., Chicago, Illinois, USA.
2 Pfizer Inc, New York, New York USA.
3 Falcon Pharmaceuticals, Ltd, Fort Worth, Texas, USA.
4 Vedco Inc., St. Joseph, Missouri, USA.
5 Kaye's Epic Compounding Pharmacy, Baltimore, Maryland, USA.