Introduction

Enucleation is indicated in rabbits to alleviate the irreversibly painful and/or blinding consequences of various ocular diseases.¹ In rabbits and other mammalian species, en bloc and subconjunctival enucleation techniques are described, with the chosen approach depending largely on surgeon preference, underlying ocular disease process, and species.^1–5 En bloc approaches like the transpalpebral and lateral techniques offer the advantage of resecting the globe and adnexa (conjunctival, eyelid margins, and third eyelid [nictitating membrane]) together in 1 piece.^2,5 However, the soft tissue dissection required for en bloc enucleations may risk intraoperative hemorrhage, which can be life-threatening, particularly in small mammals.⁴ A subconjunctival approach allows for clearer visualization of the ocular and adnexal tissues, involving initial removal of the globe followed by sequential excision of the adnexa.⁴ This technique is often advocated for rabbits on the basis of successful outcomes reported in textbooks and a small case series⁶ and the lower risk for intraoperative hemorrhage.⁴

The most significant potential source of hemorrhage encountered during a rabbit enucleation is the orbital venous sinus within the ventral and medial orbit. If transected, this large vascular structure can result in life-threatening hemorrhage, and ligation, cauterization, and tamponade techniques may not successfully halt bleeding.^5,7–9 The venous sinus anatomy in rabbits presents a unique surgical dilemma, as it is intimately associated with the Harderian gland at the deep base of the third eyelid.⁹ This proximity complicates the safe excision of the third eyelid and its secretory structures without potentially disrupting the fragile underlying venous sinus.^1,2,5 A single case report documents using an electrosurgical bipolar vessel-sealing device (LigaSure; Medtronic) to reduce hemorrhage associated with third eyelid removal during an en bloc enucleation in a rabbit.¹⁰ While this offers a promising technique to reduce the potential hemorrhage associated with third eyelid removal, many small animal and exotics practitioners may not have easy access to such devices.

In many species, enucleation without removal of secretory adnexa like the third eyelid is not recommended due to concern for associated postoperative complications such as poor postoperative healing and/or orbital cyst formation.⁵ However, there are anecdotal references to incomplete removal of the third eyelid in rabbits following a subconjunctival approach without clinically observable postoperative consequences.⁴ In other species, an evisceration technique in birds of prey involving orbital closure after leaving the third eyelid and associated glandular structures intact appears to be a viable technique with no associated long-term complications.¹¹

The objectives of this retrospective study were to review rabbit cases that underwent subconjunctival enucleation for intraoperative complications and to identify short- and long-term complications associated with this surgical procedure.

Methods

All rabbits that underwent subconjunctival enucleation surgery at the University of Wisconsin-Madison Veterinary Medical Teaching Hospital from January 1, 2014, to December 31, 2022, were identified, and medical records were compiled and reviewed.

Signalment, the reason for enucleation, the status of the contralateral eye (if applicable), treatments before enucleation, and histopathology results of the globe (if applicable) were recorded for all surgeries. Details of each surgery, including optic pedicle ligation before globe removal, number of suture layers used in site closure, and whether third eyelid removal was performed, were recorded when available. Only subconjunctival enucleations with ocular disease confined to the globe were included.

The presence or absence of intraoperative complications was recorded for all surgeries by reviewing surgery reports, medical records, and client communication notes for any mention of surgical complications. Short-term complications were recorded for each surgery, defined in this study as adverse events related to the enucleation surgery occurring ≤ 21 days from the surgery date. Long-term complications, defined as any adverse effects occurring after the 21 days, were recorded for each surgery. For all rabbits listed as surviving in the medical record system at the time of data analysis, primary care veterinarians were contacted for any additional records after the surgery was performed and owners were contacted via phone or email for an update on the surgery site. Owners were asked explicitly whether they had observed any swelling or discharge from the surgical site and/or nasal discharge from the associated nostril postoperatively or whether they would like to provide any other update about the enucleation surgery. In rabbits with unilateral enucleation, owners were also asked whether the contralateral eye had become affected by any disease processes. These responses were incorporated into the data regarding short- and long-term complications.

All enucleation surgeries performed after January 1, 2014, followed a standard subconjunctival enucleation protocol as outlined previously.⁴ In brief, a lateral canthotomy was performed, followed by a circumferential conjunctival peritomy at the level of the limbus. The conjunctiva was dissected away from the globe (Figure 1), and the rectus and oblique extraocular muscles were identified and sharply transected at their attachments to the globe. After that, the optic pedicle (optic nerve and associated vessels) and retractor bulbi muscle were transected, with or without ligation prior, and the globe was removed. The first 2 surgeries within the study period included third eyelid removal after globe removal by either clamping and sharply transecting the third eyelid or dissecting using cautery. Due to the anticipated risk of hemorrhage, the nictitating membrane was not removed in any subsequent surgeries in the study period. The superior and inferior eyelid margins were then removed via sharp dissection using tenotomy scissors. The orbital surgical site was then closed in 2 or 3 layers, depending on the amount of tissue remaining and the surgeons’ preference.

View Full Size

Subconjunctival enucleation performed in a rabbit. A—Prior to removal of the globe, the conjunctival tissue and ocular muscles were dissected from the globe. B—Postoperative appearance of the subconjunctival enucleation site.

Citation: Journal of the American Veterinary Medical Association 261, 9; 10.2460/javma.23.04.0207

• Download Figure
• Download figure as PowerPoint slide

Results

Eighteen rabbits were included in the study, 2 of which underwent bilateral enucleation, for a total of 20 enucleation surgeries. The study population included 7 male castrated, 1 female intact, and 10 female spayed rabbits. At the time of surgery, age ranged from 0.8 to 11 years old, with a median of 3.5 years old; body weight at the time of surgery ranged from 1.35 to 4.10 kg, with a median of 2.20 kg. Rabbit signalment, the reason for enucleation, treatment duration before enucleation, histopathology results when available, and whether the contralateral eye was also affected in rabbits with unilateral enucleations are summarized elsewhere (Supplementary Table S1).

Nine rabbits were suspected of having ocular manifestations of Encephalitozoon cuniculi on the basis of the presence of phacoclastic uveitis or a white-tan intraocular mass and associated uveitis at the ophthalmic examination. The median age of rabbits at the time of enucleation surgery with suspected ocular E cuniculi was 1.3 years old (range, 0.8 to 7.5 years). In 8 of 9 rabbits, 1 eye was affected and enucleated; both eyes were affected and enucleated in the remaining rabbit. The recommendation for enucleation in all rabbits with suspected E cuniculi was based on either persistent ocular discomfort despite topical anti-inflammatory medications or persistent ocular discomfort due to the inability of the owner to administer medications. Four of these 9 rabbits had histopathology results confirming ocular E cuniculi infection. One had histopathology results that were highly suspicious for E cuniculi infection, but the organism could not be visualized. Two had IgG titers ≥ 1:1,024 for E cuniculi before surgery. Of the unilateral enucleations, 1 rabbit was lost to follow-up after surgery. The remaining 7 rabbits did not develop disease in the contralateral eye, with follow-up periods ranging from 0.1 to 1.5 years (median, 0.3 years).

In 4 rabbits, enucleation was performed due to refractory glaucoma. The decision to enucleate in all rabbits was based on either persistent ocular discomfort with elevated intraocular pressure despite topical medications or the inability of the owner to medicate the rabbit. The median age of rabbits with glaucoma was 4.3 years (range, 1 to 10 years). All 4 of these rabbits had developed glaucoma in both eyes before the first enucleation surgery. One rabbit underwent bilateral enucleation. The remaining rabbits only underwent unilateral enucleation; the remaining globe was monitored and managed with topical medications as indicated, with a median follow-up period of 0.7 years (range, 0.5 to 1.4 years). Histopathology was performed in 1 of 4 rabbits but was inconclusive, showing signs of primary glaucoma with gonio-dysgenesis being the top etiologic consideration. On the basis of the ophthalmic exam before surgery, the remaining 3 of 4 rabbits did not have histopathology performed, but the etiology was suspected to be secondary to lens-induced uveitis from cataracts.

The remaining 5 rabbits underwent unilateral enucleation due to other causes, including unidentified intraocular round cell tumor, infectious keratomalacia, limbal scleral mass with glaucoma, intraocular large cell lymphoma with granulomatous inflammation, and anterior uveitis with an infected corneal ulcer. The median age of these rabbits was 8 years (range, 5 to 11 years).

Intraoperative ligation and site closure details were available for 12 of the 20 surgeries. Of these 12 surgeries, optic pedicle ligation was performed before transection using modified Miller knots or circumferential ligatures in 7 surgeries; ligation was not performed in 5 surgeries. No hemorrhage from the optic pedicle was noted in the surgery reports from either group. Eleven surgery sites were closed in 2 layers, the subcutaneous layer followed by the skin, with varying sizes of monofilament absorbable suture ranging from 3-0 to 5-0 used for the conjunctival layer and varying sizes of either monofilament absorbable or nylon nonabsorbable suture ranging from 3-0 to 4-0 for the skin. The deep layers were closed variably with continuous and interrupted suture patterns; the skin was closed in all surgeries with either simple interrupted or cruciate sutures.

The third eyelid was removed in only the first 2 of 20 surgeries. In both surgeries, sharp transection was used, facilitated using cautery in one surgery and preceded by the placement of a clamp in the other. In the latter, hemorrhage requiring prolonged digital pressure was encountered, and ligature placement around the nictitating membrane with subsequent hemostatic foam placement (Gelfoam; Pfizer) was required. No intraoperative complications were noted for the other 19 surgeries.

In 9 of 20 surgeries, information regarding use of local anesthetics was available. In 7 cases, perioperative administration of local anesthesia was recorded, including liposomal bupivacaine (Nocita) infusion into the closure layers in 3 surgeries, retrobulbar bupivacaine administration in 2 surgeries, and peribulbar bupivacaine administration in 2 surgeries. In 11 of 20 cases, information regarding the use of local anesthetics could not be retrieved due to the lack of anesthetic records available for review.

Perioperative antibiotics were administered during all surgeries using either 40 mg of ceftiofur crystalline-free acid/kg SC once or, in 1 surgery, 40,000 IU of penicillin G procaine-benzathine/kg SC on the day of surgery and again 5 days after surgery. In 2 surgeries, postoperative antibiotics were administered, including 5 mg of marbofloxacin/kg PO once daily for 14 days in 1 surgery and 30 mg of trimethoprim-sulfamethoxazole/kg PO twice daily for 14 days on the basis of culture and susceptibility results. Postoperative pain was managed in all surgeries with 1 mg of meloxicam/kg PO once daily for up to 10 days after surgery. Two rabbits were also prescribed 5 mg of gabapentin/kg PO 2 to 3 times daily for up to 10 days after surgery. All rabbits were discharged within 24 hours after surgery.

Detailed outcomes are summarized for each rabbit in the supplementary material. One rabbit was lost to follow-up after discharge from the hospital postoperatively. For rabbits with follow-up data, follow-up periods after surgeries ranged from 0.1 to 4.6 years (median, 0.7 years). Reported short- and long-term complications included mucoid to serous discharge from the surgery site in 8 of 20 surgeries. The discharge ranged from white to serous and was initially noted as early as 2 days after surgery and as late as 2.5 months after surgery (Supplementary Tables S2 and S3). The discharge persisted postoperatively for up to 10 months. Swelling of the surgery site was also noted in 2 of 20 surgeries, which was initially noted as early as 2 days after surgery and as late as 19 days after surgery. Both surgeries with swelling observed after surgery also showed postoperative discharge.

In affected rabbits, the observed postoperative discharge was not antibiotic responsive. From 1 surgical site with mucoid discharge and swelling 19 days postoperatively, a culture of the discharge grew Neisseria spp, but the growth was insufficient for susceptibility testing. The rabbit was initially treated with 40 mg of ceftiofur crystalline-free acid/kg (Excede; 200 mg/mL) for 2 injections 5 days apart with no change in the discharge. The mucoid discharge and swelling were subsequently monitored; the swelling resolved spontaneously at 3 months postoperatively, and the discharge remained intermittently until spontaneous resolution 10 months postoperatively. In a second rabbit, mucoid discharge, but no swelling, was noted 7 days postoperatively and treated empirically with 10 days of enrofloxacin with no effect on the discharge. The discharge was then monitored and spontaneously resolved at 1.5 months postoperatively. In the 5 other surgeries for which discharge was noted, spontaneous resolution occurred without antibiotic treatment. Spontaneous resolution occurred 4 days postoperatively in the other surgery for which swelling was noted.

The only other possible long-term complication noted was a mass and associated abscess formation in the caudal orbit of an enucleated eye 23 months postoperatively. This surgery site healed appropriately within 13 days of surgery and had transient serous discharge noted 2.5 months after surgery that was not noted at a vaccine visit 6 months postoperatively. The mass with associated abscess diagnosis was based on CT findings, aspiration of purulent material from the orbital swelling, and culture of Rothia nasimurium from the purulent material. The animal was euthanized due to systemic decline, and the owner declined necropsy, so histopathology of the orbit and associated pathology were not available. The CT did not show evidence of rhinitis or involvement of the cranial orbit or lacrimal duct.

Discussion

This was the first report of long-term outcomes after subconjunctival enucleations in rabbits, the majority without nictitating membrane removal. The only intraoperative complication recorded was hemorrhage during nictitating membrane removal in 1 rabbit. The only postoperative complication recorded was transient serous to mucoid discharge with or without swelling, with discharge developing in approximately 40% of the surgeries. All surgery sites ultimately healed appropriately. The observed discharge and swelling did not appear antibiotic responsive, suggesting that this complication was likely secondary to residual secretory tissue within the nictitating membrane.^4,9 The 2 surgeries with nictitating membrane removal did not exhibit any postoperative swelling or discharge. Still, the limited sample size precludes direct comparison of complication rates between surgeries with or without nictitating membrane removal.

The goal of nictitating membrane removal during enucleation is to remove the associated superficial nictitating membrane and Harderian glands to reduce postoperative secretion. However, this does not ensure the complete removal of secretory tissue.⁴ Removal of the nictitating membrane does not account for the remaining lacrimal gland in the dorsal caudal orbit and the accessory lacrimal gland that extends caudally from the nictitating membrane along the ventral orbit.⁹ Laboratory rabbits have been used as dry eye disease models, and even techniques aimed at removing all tear-producing glands have shown only a 50% reduction in Schirmer tear test results.¹² Further, there is evidence that tear-producing glands show a compensatory increase in tear production when rendered partially nonfunctional.¹³ This evidence suggests that regardless of whether the third eyelid is removed during a subconjunctival enucleation in a rabbit, tear-producing tissue will remain in the orbit postoperatively and partial removal of glands will not result in the elimination of tear production. Careful preservation of the nasolacrimal duct may help reduce postoperative swelling after enucleation without nictitating membrane removal. In a prior report of evisceration procedures in owls for which the nictitating membrane is left intact, nasolacrimal duct occlusion caused marked postoperative swelling in 1 surgery that resolved after surgery revision to remove the nasolacrimal duct occlusion.¹¹

Aside from the observed transient swelling and discharge, we cannot definitively determine whether the caudal orbital mass/abscess that developed in 1 rabbit nearly 2 years after enucleation was an actual complication of the surgery or an incidental finding. Surgical infection due to residual conjunctival tissue or communication with the nasal passage is considered unlikely on the basis of the timeline and lack of involvement of the cranial orbit. The cultured bacteria, Rothia nasimurium, is not a known inhabitant of the rabbit conjunctiva or respiratory tract but may be a component of the oral or skin flora, as it has been documented in cases of mastitis in does.^14,15 Unfortunately, without necropsy with histopathology, we cannot determine the true significance of this rabbit’s orbital pathology and associated culture results.

One notable finding of this study was that the median age of rabbits undergoing enucleation due to suspected E cuniculi ocular pathology was 1.3 years, which was lower than that of rabbits undergoing enucleation due to glaucoma or other causes (4.3 and 8 years, respectively). Ocular E cuniculi infections are most commonly thought to result from intrauterine vertical transmission resulting in higher prevalence in younger rabbits, which is supported by this retrospective study.¹⁶ To the authors’ knowledge, there are no reports regarding the prevalence of unilateral versus bilateral ocular E cuniculi infections, making it challenging to decide to enucleate an eye without known risk for contralateral disease. None of the rabbits with suspected unilateral E cuniculi ocular infections with follow-up data available developed the disease in the contralateral eye. This suggests that enucleation is a viable treatment option with a relatively low risk of contralateral disease when advanced treatment options such as phacoemulsification are not feasible.¹⁷ One rabbit exhibited bilateral ocular E cuniculi infection, but both eyes were affected before the first enucleation. While these data may help guide clinicians and owners in the decision to enucleate, it should be interpreted with caution, as a definitive histopathologic diagnosis was not achieved in most rabbits in this study.

After E cuniculi infection, glaucoma was the second most common reason for enucleation in this retrospective study. Glaucoma in rabbits has historically been most commonly attributed to congenital primary glaucoma, which is typically diagnosed early in life.¹⁸ The age range of rabbits undergoing enucleation due to glaucoma in our study was 1 to 10 years, which is similar to the age range of rabbits diagnosed with glaucoma in a recent review¹⁹ from a different teaching hospital (8 months to 12 years). This review highlighted the fact that primary congenital glaucoma may be a less common cause of glaucoma in pet rabbits than previously assumed.²⁰ Unfortunately, on the basis of data available to us, we cannot comment on the definitive cause of glaucoma in the cases included in our study. This study in conjunction with the aforementioned recent review on glaucoma in pet rabbits highlights the knowledge gaps in the etiology and subsequently best treatment options for pet rabbits with glaucoma.

This study had several limitations, including a relatively low sample size, although similar to prior studies in other species.^11,20 A more robust sample size with more rabbits undergoing nictitating membrane removal may help further characterize the intraoperative risk of nictitating membrane removal and comparison of complication rates. Additionally, this was a single-institution study with a fairly standardized surgical and postoperative management plan; changes to these factors may affect the complication rate when surgery is performed at different institutions. For example, the length of surgery and appropriate use of perioperative antibiotics are just 2 known factors that affect the rate of surgical site infections.^21–23 Detailed anesthetic records were not available for all surgeries; rabbits are prone to intraoperative hypotension, which can reduce intraoperative hemorrhage, thus decreasing our incidence of intraoperative complications.^24,25 Further, follow-up information was obtained via veterinary records or owner reporting, with owner reporting potentially susceptible to recall bias. Histopathology of the surgery sites and associated orbits would be the ideal way to evaluate the outcome. Still, it could not be performed in most included rabbits due to either owners declining necropsy or the rabbit being alive at the time of data collection.

In conclusion, subconjunctival enucleation in rabbits with or without nictitating membrane removal appears to be a viable surgical approach with a low rate of postoperative complications and favorable long-term outcomes of up to 56 months postoperatively in this study. The postoperative complications observed were generally transient and self-limiting. There were no intraoperative complications in enucleations without nictitating membrane removal; hemorrhage occurred in 1 enucleation with nictitating membrane removal. Further investigation should be done to better evaluate the effect of nictitating membrane removal on complications and outcomes and other interinstitutional factors that may affect outcomes in these cases.

Supplementary Materials

Supplementary materials are posted online at the journal website: avmajournals.avma.org