Introduction

In cats, malignant tumors of the eyelid can be locally controlled by surgery and/or radiotherapy. If surgery is an option, a wide excision involving a variable portion of the eyelid and surrounding skin and eye enucleation/exenteration are performed. If the latter is not feasible for any reason, tumor excision without enucleation/evisceration and reconstruction is indicated.¹ Ideally, the reconstruction should create a hairless, smooth-lid margin to avoid trauma to the cornea and to preserve movement of the eyelid.² Eyelid tumors are less frequently encountered in cats than in dogs; however, the incidence of malignancy is higher in cats.¹ There are only a few reports describing feline upper eyelid (UE) reconstruction, especially in cases of wide defects, and no gold standard has been established.^3–8 In fact, all the described techniques carry the risk of complications,^3–8 and sometimes, more than one surgery is necessary.^1,3,4 Surgery using a mucocutaneous subdermal plexus flap (lip-to-lid flap) has been reported for a cat after a wide UE removal.⁸ The lip-to-lid technique provides a flap lined with mucosa and a hairless margin; however, it can be too narrow to reconstruct a large defect of the UE, especially along the medial canthus. The use of a superficial temporal artery axial pattern flap has also been proposed, but the close proximity to the excisional area can impede its use for wide surgical defects of the UE.⁹ In addition, a flap based on a branch of the facial artery can be used, especially for lateral defects.¹⁰

It has been reported that for cases of wide excision of the eyelid, the conjunctiva needs to be reconstructed to restore a physiologic coverage of the eyelid and prevent irritation/damage to the cornea.¹ In dogs, substitution of the conjunctiva can be performed with a two-step procedure using the conjunctiva from the lower eyelid, or it can involve an oral mucosa–free graft, the nictitans conjunctiva, or by creating the flap from tissue that already has a mucosal lining, such as the lip-to-lid flap.^8,11,12

A caudal auricular axial pattern flap (CAAPF) procedure has been used in cats to repair defects of the ear and the frontotemporal and orbital area, even after eye enucleation/exenteration.^13,14 The CAAPF can be used for large defects involving the entire UE, but reports on its use for the repair of such defects are lacking. The purpose of this paper is to describe the use of CAAPF for single-step UE reconstruction in three cats following en bloc tumor excision.

Case Report

Three cases involving spayed female domestic shorthairs are included in this report. The ages of the cats were 9 (cat 1) and 11 (cats 2 and 3) yr. Preoperative cytology identified two squamous cell carcinomas (SCCs; cats 1 and 3; Figure 1A) and one soft-tissue sarcoma (cat 2; Figure 2A). The tumor sizes were 2.5 × 3, 1.4 × 0.5, and 3 × 3.5 cm, respectively. The complete blood cell counts and biochemistry panels were within normal limits for all animals; cat 3 was feline immunodeficiency virus positive. The tumor clinical staging consisted of regional lymph node fine-needle aspiration and cytological examination, three radiographic views of the thorax, and an abdominal ultrasound reviewed by an experienced radiologist. No metastases were detected in any cat. Because of the site and size of the tumors, wide en bloc excision with exenteration, or marginal tumor excision with adjuvant radiation therapy, was recommended in all cases; however, these treatments were declined by the owners. Therefore, an alternative option of en bloc excision without exenteration was offered, with the understanding that complete excision of the tumor might not be achieved. Moreover, further possible complications associated with complete UE reconstruction, such as trichiasis, decreased ability to blink, decreased tear production, development of corneal ulcers, and a possible altered cosmetic appearance, were discussed with the owners in advance.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Premedication was provided by intramuscular injection of dexmedetomidine (0.005 mg/kg), ketamine (3 mg/kg), and methadone (0.2 mg/kg). After placement of an IV catheter, induction of anesthesia was obtained with propofol to effect (2–4 mg/kg IV), and after orotracheal intubation, anesthesia was maintained with isoflurane in a mixture of air and oxygen. Analgesia was maintained throughout the surgical procedure by a constant rate infusion of fentanyl (0.005–0.010 mg/kg/hr). Cefazolin^a (20 mg/kg IV) was administered at the induction of anesthesia. Cats were clipped on the head and neck, including the ipsilateral scapular region; laid in lateral recumbency with the tumor side up; and prepared for surgery, taking care to avoid direct contact between the antiseptics and cornea. The conjunctiva and the cornea were irrigated with sterile NaCl 0.9% and 0.2% diluted povidone iodine solution, and a cover of sterile eye gel^b was applied over the corneal surface to prevent corneal desiccation during surgery. The tumor was excised, together with a minimum of 1 cm of grossly healthy tissue surrounding the mass and the complete thickness of the eyelid up to the bulbar conjunctiva. The lateral canthus was excised in all cats; the medial canthus was only partially excised in cats 2 and 3 and completely removed in cat 1 (Figures 1B, 2B).

After changing surgical instruments and gloves, CAAPF was prepared as previously described.¹⁴ Briefly, two parallel skin incisions were performed, caudal to the scutiform cartilage of the pinna and along the dorsolateral neck toward the cranial margin of the scapula. The length of the flap varied from 10 to 12 cm in length and 4 to 5 cm in width. The bridging incision, allowing caudocranial rotation of the flap, was medial to the ipsilateral ear in cat 1 (Figure 1C) and ventral in cats 2 (Figure 2C) and 3. The subcutis and skin of the donor site were sutured with 3/0 absorbable monofilament material^c and/or stainless-steel staples^d. In an attempt to avoid trauma to the cornea, the distal free margin of the flap was sutured to the skin of the medial canthus and to the lower eyelid with figure-of-eight sutures with very short tags, using 4/0 absorbable sutures^e in cat 1 (Figure 1C), and with 5 cm long tags sutured to the temporal skin in cat 2 using 3/0 monofilament nonabsorbable sutures^f (Figure 2C).¹ In the second cat, the suture tags were left long and sutured away from the eye to prevent corneal contact and damage. The remaining part of the distal margin of the flap was left unsutured. To adapt the distal margin of the flap to the defect, a small central skin fold was excised, and its margins were sutured perpendicularly to the flap edge in cat 1 (Figure 1C). In cat 2, the distal margin of the flap was beveled craniocaudally, creating a scarred surface in an attempt to avoid hair regrowth (Figure 2C). In cat 3, the medial and lateral margins of the flap were sutured to the excision margins, using mattress pattern sutures (3/0 absorbable monofilament), and its free edge was sutured to a narrow strip of 2 mm of preserved conjunctiva, using mattress pattern sutures (3/0 absorbable monofilament) with short tags. An active suction drain was placed at the donor site in cat 2 and a passive drain in cat 3, and they were removed after 2 and 7 days, respectively.

The cats were hospitalized for 3–7 days postoperatively and wore an Elizabethan collar until complete healing occurred. The postoperative medications consisted of 0.1 (first day) and 0.05 (following days) mg/kg meloxicam^g q 24 hr per os for 3 days and 0.015 mg/kg buprenorphine^h IV q 6–8 hr for 3 days. Artificial tearsⁱ and topical antibiotic ointment^j were locally applied four times a day until healing. The cats were rechecked 3 and 12 days after discharge, and then every 1–2 mo by the referring veterinarian for the first 6 mo after surgery, followed by a minimum of once every 6 mo.

In cases 1 and 3, a completely excised SCC was histologically confirmed. A noncapsulated epithelial neoplasia, with a diffuse squamous differentiation cells, was present in both cases. Cells were cuboid-to-polygonal shaped with eosinophilic cytoplasm, vesicular nuclei, and 1–2 nucleoli. Anisocytosis and anisokaryosis were moderate, with four mitoses per high-power field in case 1 and eight mitoses per high-power field in case 3. In case 2, a malignant schwannoma (MS), infiltrating dermis, subcutis, and muscle layers of the UE, with neoplastic cells extending over the surgical margins, was the final diagnosis. Sheets or fascicles of spindle cells with elongated nuclei were organized in row or palisades, and other portions of the tumor showed a looser arrangement of spindle cells in a fibrovascular stroma. Mitoses were rare, and areas of necrosis and neutrophilic infiltration were present. Adjuvant radiotherapy was recommended for case 2 but was declined by the owner.

Healing occurred in all cases; early complications consisted of a transient edema and chemosis lasting 5–7 days in all the cats. All three cats also experienced epiphora during the first 15–20 days. A slight overexposure of the bulbar conjunctiva occurred in cat 3. After 10 and 20 days, mild exudation was present in cats 1 and 3, respectively, and superficial corneal ulcers were diagnosed in both cats using a fluorescein test. The Schirmer tear test was normal (14 mm/min, range 14 ± 6 mm/min) in cat 1 after 10 postoperative days. The ulcers healed using antibiotic ointmentⁱ (two drops every 6–8 hr for 10 days), and autologous serum (one drop every 6 hr) was also applied in case 3. Trichiasis occurred in cat 1 and was managed with periodical hair trimming. The procedure was stopped when the owner realized that the longer hair was well tolerated (as it was growing in the opposite direction relative to the cornea) and did not interfere with eye function (Figure 1D). Partial blinking was still possible, thanks to the movement of the lower and third eyelids. Eye asymmetry was evident in cats 2 and 3, owing to the ventral tensing of the new lid and to decreased eye coverage, respectively.

A histologically confirmed in situ SCC developed 807 days later on the temporal area of the same side and was successfully excised in cat 1; this cat died of an alimentary lymphoma 6 yr later. In case 2, a local recurrence of the MS developed 350 days after the first excision; however, the owners declined any further treatment, and the cat was lost to follow-up. Cat 3 experienced a tumor recurrence after 380 days, and at that time, exenteration was performed; this cat died of renal failure 763 days after the first surgery without any further tumor recurrence.

Discussion

The UE is a challenging area to deal with surgically because of the need for the preservation of lid function. In the case of large malignant tumors, eye enucleation/exenteration is often indicated to achieve tumor-free histological margins. However, in selected cases, and when the owners are opposed to exenteration, en bloc eyelid excision and reconstruction may be considered.

The lip-to-lid flap, superficial temporalis flap, and random local advancement, transposition, or rotation flaps based on the subdermal plexus are proposed techniques for the repair of eyelid defects.^7,8,15,16 However, these flaps were not considered wide enough to repair the defects in the three reported cats. In fact, advancement flaps need to be twice the length of the defect; other potential limitations in the repair of eyelid defects are associated with contraction during healing and the possible distortion of the eyelid margin. Rotational flaps can restore no more than 60% of the length of the eyelid, and local subdermal flaps have limitations with regard to the extension of the flap. Usually, the width should be at least 50% of the length of the flap, and the apex of the flap should be smaller than the base.^1,17 The superficial temporalis flap was not suitable for the three cats because the area of excision would overlap with the defected area in both cats 1 and 2.⁹ Other flaps, such as the cross-lid flap and the Bucket handle techniques, have several disadvantages; they require two surgical procedures, cause impaired vision during the period between the two procedures, and provide insufficient tissue to cover large defects, such as in the cats described in this report.¹ A suitable axial pattern flap for these cats would have been based on the facial artery, as it is closer to the defect created; however, the surgeon’s personal choice in these cases was the CAAPF, which had been reported to provide a large amount of vascularized skin and to be a suitable procedure for covering defects created after orbital exenteration in cats.^10,14 This flap was able to restore the eye coverage in all three cats, with a slight bulbar overexposure only in one cat (cat 3). Moreover, the proposed reconstruction of the UE required one single-stage procedure and provided a wide flap that was easily adaptable to the defect; however, short- and long-term complications were encountered. In the short-term in particular, corneal ulceration, trichiasis, and a reduced ability to blink were recorded; in the long term, apart from tumor recurrences, there was a slight asymmetry of the eyelid and subsequent overexposure of the bulbus in cat 3. No complications occurred at any of the donor sites.

It was necessary to suture only the margins of the flaps in place (which became the new upper portion of the palpebral fissures) because of the lack of adequate conjunctival tissue left after the en bloc removal of the tumor. Conversely, in cat 3, a small portion of conjunctiva was spared to cover the inner part of the reconstructed eyelid. However, it resulted in a slight overexposure of the bulbar region, possibly as a result of scar tissue retraction. No attempt was made to cover the skin flap with conjunctiva or oral mucosa in cats 1 and 2. Stades reported that adjacent conjunctiva may epithelialize the deep surface of the apposed skin.¹⁸ In both cats 1 and 3, who underwent a second surgery at 807 and 380 postoperative days, respectively, the reconstructed eyelid was covered by tissue macroscopically judged to be conjunctiva.¹⁸ Therefore, the role of the skin that is in direct contact with the cornea in the development of the corneal ulcer of cat 1 cannot be established. Corneal ulcers developed in two cats (cats 1 and 3), possibly as a result of the short suture tags in contact with the cornea, exudation and crust at the medial canthus, as well as overexposure of the globe in cat 3 and lack of conjunctiva coverage of the new UE in cat 1. Moreover, we cannot rule out decreased movement of the UE or poor lubrication as inciting causes, even though the Schirmer tear test was normal for cat 1. In cat 2, figure-of-eight sutures with long tags were applied and ulcer formation was avoided, even though the skin was not covered by conjunctiva.¹ To prevent corneal ulceration, attention should be paid to placing suture tags that do not touch the cornea even during postoperative swelling of the surgical site. It is likely that long suture tags sutured away from the cornea can avoid cornea abrasion. Because of the frequency of corneal ulceration (two of the three cats), close monitoring in the postoperative period is required to detect this complication as soon as possible and provide adequate treatment.

Flap tip necrosis is the most frequent surgical complication reported for CAAPF.^14,19 However, it did not occur for any of the cats in this report. The complete healing of the flaps that occurred in these three cats could be related to the shorter flap length required to reconstruct the eyelid compared with the length necessary to repair a wound after exenteration. In the present cases, the flap length never reached the spina scapularis, which is considered the maximum length possible for this flap.¹⁴

Regarding the en bloc resection of the UE, there were concerns about epiphora, lacrimal gland loss, trichiasis, and a lack of normal lid movement. In cats 1 and 2, the upper nasolacrimal punctum was excised; however, this did not result in long-term epiphora, as already reported by Hagard.⁶ Cat 1 developed trichiasis, which was initially managed by hair trimming. However, because of the growth of the hair away from the eye, this was well tolerated by the cat and was not a major complication in the long term. In an attempt to avoid trichiasis, the margin of the flap may also be tapered to prevent hair growth close to the cornea, as in case 2.

Blinking was affected in all the cases, as the UE is usually more movable than the lower eyelid and is the major structure responsible for tear distribution. However, after an initial period, during which artificial tears were applied several times per day, both the lower lid and third eyelid seemed to compensate, maintaining adequate lubrication. In fact, during the blink reflex, the bulbus retracted and the third and lower eyelid moved and most likely distributed the tear film over the cornea. However, the Schirmer tear test should always be performed, both preoperatively and postoperatively, to detect tear production deficits. Unfortunately, it was performed only in cat 1.

The position of the bridging incision during the flap rotation was primarily the surgeons’ preference, with no difference in terms of CAAPF coverage or healing. The dorsal rotation, however, enabled easier modeling of the palpebral fissure and a better cosmetic result, in comparison with the ventral rotation of the flap. When ventral rotation is performed, the possibility that the weight of the skin could distort the shape of the palpebral fissure should be considered. The position of the bridging incision should be evaluated based on the extension and location of the defect.

The eye was preserved in all cases, and after an initial period of intensive care to prevent or treat complications, the reconstructed UE provided coverage to the eye in the long term. No long-term complications were observed in terms of eye function except for tumor recurrences. The cosmetic result was considered satisfactory, even when the hair in the region was longer and grew in the opposite direction to the contralateral eyelid hair. This was not a problem for the animals or the owners; however, its possibility should always be discussed with the owners before surgery.

The tumors found in the three cats were all malignant and locally invasive with a low rate of distant metastasis, which usually occurs late in the course of the disease. However, when these lesions are left untreated, their locally aggressive behavior often leads to euthanasia.²⁰ The en bloc UE removal and reconstruction controlled the disease locally, preserving the eye for at least 350 days, even in cat 2, who had an incomplete tumor excision. The MS recurrence in cat 2 was expected following the histological report of tumor infiltrated margins and the lack of further adjuvant therapies. In fact, recurrences are frequently reported in these tumor types, especially following incomplete excision.²⁰ Conversely, in cat 3, the recurrence of the SCC after 380 days was unexpected. In this case, the presence of neoplastic cells beyond the excisional margins may have been missed in the histological evaluation as a result of technical limitations. It is well known that some tumors tend to recur even after complete removal has been histologically evaluated.²¹ In cat 1, the development of an in situ SCC in the temporal area 807 days after surgery could have been either a new tumor or a recurrence of the first neoplasm. Although there is no gold standard by which to answer the question of recurrence versus de novo tumor development, the length of time between tumor occurrences, the location of the second tumor being far away from the first one, white fur color, and continuous exposure to ultraviolet light suggest the development of a second de novo tumor of the same histotype rather than a tumor recurrence.

Conclusion

In the case of large eyelid malignancy, wide local excision with exenteration may provide more effective local tumor control than en bloc excision of the UE. However, the en bloc excision of the UE and reconstruction with CAAPF may be considered for small tumors or when the owner refuses exenteration, although the need for appropriate and prolonged postoperative eye care should be clearly communicated to the owner. CAAPF, as a one-step procedure, might be an option for the reconstruction of the UE region in selected cases, when the aim is to maintain good eye coverage, function, and an acceptable cosmetic appearance.