Abstract
A case of cutaneous haemangiosarcoma of the left lower eyelid in a 15-year-old white domestic shorthair cat is reported. A protuberant red mass occupying one-third of the lower eyelid margin length was present. Intermittent haemorrhage occurred from the mass surface. Surgical biopsy had revealed a locally invasive tumour composed of numerous irregular blood-filled spaces lined by a single layer of plump endothelial cells and separated by thin fibrous septa. Mitotic activity was rare and the appearance was consistent with a low-grade haemangiosarcoma. The mass continued to enlarge and referral was sought. Due to financial constraints and the owner's wish for a single procedure, enucleation with an axial pattern flap based on the superficial temporal artery was undertaken. Histopathology of the excised tissue confirmed the presence of a well-differentiated haemangiosarcoma. Tumour-free excisional margins were confirmed, the surgical area healed uneventfully, and there has been no recurrence during the subsequent 16 months.
A 15-year-old female spayed white domestic shorthair cat was presented with a bleeding red fleshy mass on the lower eyelid of the left eye. The mass occupied one-third of the eyelid margin length. The owner had noted the mass 2 weeks previously; there had been recent intermittent haemorrhage from its apical surface. At the time of initial clinical examination, the cat weighed 2.95 kg. A small surgical sliver biopsy was consistent with a low-grade haemangiosarcoma. The mass continued to enlarge rapidly and a referral was sought for further investigation and treatment.
At referral, 5 months after the owner noted the eyelid mass, ophthalmic examination revealed a friable red fleshy mass on the left lower eyelid (Fig. 1). The apical surface was ulcerated and bleeding, and dried blood was present over the left side of the cat's face. The mass extended ventrally from the middle of the left lower eyelid margin for approximately 7 mm, and both laterally and medially along the eyelid margins, occupying approximately two-thirds of the eyelid margin length. The mass also extended onto the palpebral conjunctiva for approximately 5 mm. The remainder of the ophthalmic examination was unremarkable. The cat was underweight, and having weight since initially examined by the referring veterinary surgeon (bodyweight 2.75 kg) and had an unkempt coat. Abdominal palpation revealed no abnormalities and there was no evidence of a peripheral lymphadenopathy. A palpable unilateral right-sided goitre of approximately 1 cm×6 mm was present. Heart rate was recorded as 190 bpm with no audible murmur, systolic blood pressure was recorded as 162 mmHg (mean of eight readings, Doppler). Results of serum biochemistry and haematology panels were within normal limits, with exception of alanine aminotransferase (459 U/l, normal range 5–60 U/l), alkaline phosphatase 162 U/l (<60 U/l), urea 10.9 mmol/l (2.5–9.9 mmol/l). Serum thyroxine was 220 nmol/l (normal range 10–55 nmol/l). The owner reported no signs of polyuria, polydipsia or polyphagia but felt that the cat had appeared to lose weight over the last few months. Thoracic and abdominal radiographs were within normal limits with no evidence of metastatic disease. The owner declined abdominal and cardiac ultrasonography, and fine needle aspiration cytology of the regional lymph node on financial grounds. Treatment for hyperthyroidism was initiated with 5 mg methimazole (Felimazole; Arnolds) bid.
Appearance of mass at presentation.
The owner was anxious to avoid repeat surgery and in order to achieve tumour-free margins, enucleation with a wide surgical excision and an advancement skin graft based on the superficial temporal artery was proposed. The cat was anaesthetised and transconjunctival enucleation undertaken. The mass was then excised with a gross margin of 1 cm, and using a skin graft harvested from the superficial temporal region, the skin defect was closed (Fig. 2). Histological examination of the mass revealed a well-defined vascular neoplasm located within the dermis of the eyelid, with elevation and erosion of the overlying epidermis (Fig. 3). The neoplasm consisted of multiple capillary-sized or smaller blood-filled spaces lined by swollen and pleomorphic endothelial cells and supported by fine interconnected collagenous septa (Fig. 3). Constituent endothelial cells demonstrated mild to moderate nuclear atypia and scattered mitoses (less than one per high power field). The smaller vascular channels sometimes lacked clear lumina and presented as slit-like spaces (Fig. 3). The margins of the neoplasm were expansible to mildly infiltrative, with no evidence of encapsulation. The excisional margins were complete with tumour-free margins of at least 5 mm in all sections.
Post-surgical appearance.
Histopathology of mass ×200. Note multiple capillary channels or smaller slit-like spaces containing red blood cells.
The surgical area healed well postoperatively. The hyperthyroidism remained controlled medically and the cat gained weight (body weight=3.15 kg) after methimazole treatment.
Haemangiosarcoma is a malignant neoplasm of vascular endothelial origin. Reports of haemangiosarcoma in the cat are rare, accounting for less than 2% of all non-haematopoietic neoplasms (Patnaik and Liu 1977, Scavelli et al 1985, Miller et al 1991, Kraje et al 1999). Both cutaneous (skin and subcuticular tissues) and visceral (mesentery, spleen, liver, mediastinum, myocardium, lung and bone) haemangiosarcomas have been reported in cats (Patnaik and Liu 1977, Scavelli et al 1985, Miller et al 1992). In contrast to visceral haemangiosarcoma, feline cutaneous haemangiosarcoma has generally been reported to metastasise rarely, although local recurrence can be a significant problem (Patnaik and Liu 1977, Miller et al 1991, 1992).
Cutaneous haemangiosarcoma has been reported most frequently on the head (Patnaik and Liu 1977, Miller et al 1991, 1992), and a link with chronic solar irradiation postulated (Miller et al 1992). A similar conclusion was also drawn in a study of canine cutaneous haemangiosarcoma, where solar elastosis was sometimes seen in association with haemangiosarcoma (Dorn 1976). In addition, that study identified a marked predilection for canine cutaneous haemangiosarcoma in lightly pigmented or glaborous skin (particularly of the ventrum), and for short-haired dogs. The cat in this case was both white and short haired. White cats have been shown to be at greater risk of palpebral squamous cell carcinoma and it is possible that similar actinic preneoplastic changes occurred in this case which subsequently progressed to haemangiosarcoma (Dorn et al 1971).
Palpebral neoplasia in cats is not as common as it is in dogs but its prevalence increases with advancing age (Dorn 1976, Williams et al 1981, McLaughlin et al 1993). A review of the literature revealed a total of 17 different feline palpebral tumour types with squamous cell carcinoma being the most common, accounting for 65% of the total in one study (Williams et al 1981, McLaughlin et al 1993).
A single report of feline palpebral haemangiosarcoma exists as part of a series of cutaneous haemangiosarcoma cases (McAbee et al 2005). This case received no treatment and was euthanased 2 months later due to tumour enlargement and haemorrhage.
Haemangiosarcoma of the third eyelid has also been reported in a cat. Surgical excision with cryotherapy had a favourable outcome (no recurrence after 7 months) (Multari et al 2002, Liapis 2004).
One retrospective study involving 31 cases of feline haemangiosarcoma reported that 60% of cutaneous and subcutaneous cases had local recurrence but none had evidence of metastasis (Scavelli et al 1985). The mean time to local recurrence was 16 weeks with a range from 2 to 32 weeks. In contrast, these authors reported a high metastasis rate (60%) for cases of abdominal haemangiosarcoma. The mean age at diagnosis of haemangiosarcoma (all sites) in this study was 9.6 years (range 4–15 years) (Scavelli et al 1985). Another study of 15 cats reported a median age of 12.5 years for cutaneous haemangiosarcoma with a predilection for unpigmented skin of the head (in particular the pinnae) and for male cats (Miller et al 1992). Although most studies suggest a low metastatic rate for cutaneous haemangiosarcoma in cats, one study has reported metastases in five cats (Kraje et al 1999).
Haemangiosarcoma is a more common tumour in dogs than cats and has been reported in many visceral sites (Pearson and Head 1976, Oksanen 1978, Brown et al 1985, Ward et al 1994). Metastasis from visceral sites is generally rapid and widespread. Dogs have a poor prognosis overall, although the prevalence of metastasis from cutaneous or subcutaneous sites is similarly lower than from visceral sites (Dorn 1976, Waters et al 1988, McLaughlin et al 1993, Smith 2003). However, confusion can exist – canine cutaneous haemangiosarcomas are reported uncommonly as primary tumours, but visceral haemangiosarcomas may readily metastasise to the skin and thus many dogs with cutaneous manifestations have metastatic disease from a visceral primary. The diagnosis of haemangiosarcoma affecting the skin, which would include metastases to that site, is therefore often associated with a poor prognosis (Dorn 1976, Oksanen 1978, Waters et al 1988).
Histological examination and staging was not helpful in predicting clinical outcome of feline cutaneous haemangiosarcomas in a study of 15 cases of feline cutaneous vascular neoplasia (Miller et al 1992). McAbee et al (2005) reported longer survival times in cases with deep tissue involvement (subcutaneous and muscle involvement) compared to those with dermal tumours. In the case presented here the haemangiosarcoma was confined to the dermis with no invasion of underlying structures (orbital rim or musculature).
In view of the reported low metastatic risk but high frequency of local recurrence for these tumours in cats, a wide surgical excision was recommended. Due to the location and size of the mass in this case wide excision was suggested by either lip to lid blepharoplasty (Pavletic technique), the two stage Mustarde or bridge-flap blepharoplasty techniques, or by enucleation with an axial pattern flap based on the superficial temporal artery to fill the defect (Doherty 1973, Pavletic et al 1982, Fahie and Smith 1997). The owner desired a single procedure with the greatest excisional margins possible. For this reason enucleation with a superficial temporal artery axial pattern flap was undertaken.
Axial pattern flaps are pedicle grafts that incorporate a direct cutaneous artery and vein to their base. The terminal branches of these vessels supply blood to the subdermal plexus. Axial pattern flaps have better perfusion compared with pedicle grafts for which circulation is derived from the subdermal plexus alone (Pavletic 1990). The axial pattern flap used in this case was harvested from the frontal region and extended to the middle of the dorsal orbital rim of the contralateral (right) eye, providing adequate tissue to reconstruct the orbital region of the left eye, as well as primary closure of the donor site without excessive tension (Hargis et al 1992).
The cat reported here remains tumour free 16 months after excisional surgery. The prolonged survival supports the hypothesis that the eyelid mass was a primary cutaneous tumour rather than metastatic spread from visceral disease. To the authors' knowledge, this is the first detailed report of treatment of feline palpebral haemangiosarcoma, and it demonstrates the potential for a good prognosis with aggressive surgery.