A Retrospective Study of Eyelid Tumors from 43 Cats

Abstract

Submissions to the University of Tennessee pathology service from June 1999 to June 2008 were searched for feline cases of tumors involving the eyelids or nictitans. Forty-three tumors were identified. The average age at diagnosis was 10.4 years. Significantly more males than females had eyelid tumors. There were 12 squamous cell carcinomas (SCCs), 11 mast cell tumors (MCTs), 6 hemangiosarcomas (HSAs), 4 adenocarcinomas (ACAs), 3 peripheral nerve sheath tumors (PNSTs), 3 lymphomas, 3 apocrine hidrocystomas (AHCs), and 2 hemangiomas. Cats with MCTs were significantly younger than cats with all other tumor types combined. In contrast, cats with SCCs were significantly older than cats with other tumor types. The HSAs and SCCs were significantly more likely than other tumors to occur in nonpigmented areas. The MCTs, HSAs, AHCs, and hemangiomas did not recur after surgical excision. In contrast, the lymphomas, ACAs, SCCs, and PNSTs frequently recurred and/or resulted in death or euthanasia of the cat. The SCCs were significantly more likely to recur than the MCTs. The average survival time for cats with SCCs was 7.4 months. Although eyelid MCTs have been reported in cats, the prevalence in this study is much higher than previously described.

Tumors and tumor-like lesions of the eyelids are less common in cats than in dogs but more likely to be malignant. ²⁶ Squamous cell carcinoma is the most common eyelid neoplasm reported in cats. ^7,17 The last comprehensive study of the prevalence of feline eyelid tumors published in 1993 from the Veterinary Medical Data Program (VMDP) and the Purdue Comparative Oncology Program (PCOP) ¹⁷ listed squamous cell carcinomas (SCCs), squamous papillomas, uncategorized carcinomas, basal cell carcinomas, fibromas, fibrosarcomas, adenomas, cystadenomas, adenocarcinomas (ACAs), lymphomas, histiocytomas, mast cell tumors (MCTs), hemangiomas, hemangiosarcomas (HSAs), melanomas, neurofibromas, and trichoepitheliomas. ¹⁷ More recently, there have been isolated case reports and case series describing apocrine hidrocystomas (AHCs), hemangiomas, HSAs, lymphomas, and peripheral nerve sheath tumors (PNSTs) in the eyelids of cats. ^{3,5,12–14,20,22} The current report describes the feline eyelid tumors submitted to the University of Tennessee between June 1999 and June 2008.

Materials and Methods

Case identification

Necropsy and biopsy submissions to the University of Tennessee from area practices or the Veterinary Teaching Hospital were searched from June 1999 to June 2008 for feline tumors of the eyelids or nictitans. Only those tumors that specifically involved the lids were included; lesions described as near the eyelid or periocular were excluded. Available medical records were reviewed and information about the signalment, location, recurrence, surgical description, and ancillary diagnostic tests was tabulated. Follow-up information from the owner or veterinarian was used to determine final outcome.

A single pathologist (KMN) reviewed the hematoxylin and eosin–stained sections to confirm the diagnosis for each case. Nondiagnostic samples were excluded. The histologic classifications were based on published definitions of the tumors. Briefly, SCCs were defined by cords and nests of neoplastic epithelial cells that demonstrated progression from peripheral basal cells to central keratinized cells, with variable formation of keratin pearls (Fig. 1). ^8,25 Features of malignancy included anisokaryosis, large prominent nucleoli, frequent and often atypical mitotic figures, and invasion of the surrounding tissues. ^8,25 Mast cell tumors were composed of sheets and cords of uniform round cells with discrete cytoplasmic margins and variable numbers of basophilic cytoplasmic granules (Fig. 2). ¹⁰ HSAs were characterized by plump endothelial cells aligned on delicate collagenous trabeculae that formed an anastomosing meshwork of variably sized and often incompletely formed blood-filled channels (Fig. 3). ^11,25 Features of malignancy included anisokaryosis, increased nuclear to cytoplasmic ratio, prominent nucleoli, and evidence of invasion into surrounding tissues. ^11,25 ACAs were defined by the presence of invasive nests and cords of pleomorphic epithelial cells with acinar formations (Fig. 4). ²⁵ PNSTs had spindle cells arranged in sheets and fascicles (Fig. 5). ¹³ Lymphomas were characterized by infiltrative sheets of monomorphic lymphocytes with prominent nucleoli and frequent mitoses (Fig. 6). ^9,25 AHCs were cystic structures that originated multifocally from palpebral apocrine sweat glands and were lined by 1 or more layers of cuboidal to columnar epithelium (Fig. 7). These tumors frequently had short papillary projections extending into the lumen. ^3,24 In contrast to the HSA, hemangiomas consistently formed well-defined, blood-filled vascular spaces lined by well-differentiated endothelial cells (Fig. 8). ^11,25

Fig. 1.

Squamous cell carcinoma, No. 10, HE, ×400.

Fig. 2.

Mast cell tumor, No. 19, HE, ×400.

Fig. 3.

Hemangiosarcoma, No. 25, HE, ×400.

Fig. 4.

Adenocarcinoma, No. 30, HE, ×400.

Fig. 5.

Peripheral nerve sheath tumor, No. 36, HE, ×400.

Fig. 6.

Lymphoma, No. 39, HE, ×400.

Fig. 7.

Apocrine hidrocystoma, No. 41, HE, ×400.

Fig. 8.

Hemangioma, No. 43, HE, ×400.

When available, the adjacent non-neoplastic epithelium was categorized as pigmented or nonpigmented. Histochemistry (toluidine blue) and immunohistochemistry were performed as needed to further characterize the neoplastic cells.

Immunohistochemistry

Antibodies, concentrations, antigen retrieval methods, blocking methods, and incubation times are listed in Table 1. For chromogen detection, all slides were incubated for 30 minutes in either the mouse or rabbit EnVision+ reagent (Dako, Carpinteria, CA) before a 10-minute detection step using diaminobenizidine, as appropriate. All incubations were performed at room temperature.

Table 1.

Primary antibodies and techniques used for immunohistochemistry.

Primary Antibody	Source	Antigen Retrieval	3% Hydrogen Peroxide Block	Serum-Free Protein Block (Dako)	Dilution	Time (minutes)
CD3 (monoclonal mouse anti-human clone F7.2.38)	Dako	25 minutes in a steamer at 95°C with EDTA (pH 9); 20-minute cool down	5 minutes	5 minutes	1:1,000	30
CD79a (monoclonal mouse anti-human clone HM57)	Dako	25 minutes in a steamer at 95°C with EDTA (pH 9); 20-minute cool down	5 minutes	20 minutes	1:1,000	30
CD18 (mouse anti-feline, Fe3. 9F2-IgG1)	Dr. Peter Moore, University of California, Davis	5 minutes in proteinase K	5 minutes	None	1:100	30
CD117 (c-kit; polyclonal rabbit anti-human)	Dako	25 minutes in a steamer at 95°C with EDTA (pH 9); 20-minute cool down	5 minutes	5 minutes	1:500	60
S100 (polyclonal rabbit anti-human)	Dako	Non	5 minutes	5 minutes	1:2,000	30
SMA (mouse anti-human clone 1A4)	Dako	25 minutes in a steamer at 95°C with EDTA (pH 9); 20-minute cool down	5 minutes	5 minutes	1:1,000	30

Data analysis

Continuous data were tested for fit to a normal distribution using the W statistic of Shapiro-Wilk. Data are expressed as mean ± SD or mean ± SEM. Student's t-test was used to compare age among cats with SCCs and MCTs and the presence of pigment in the epithelium among cats with SCCs or HSAs with all other cats. Other tumor types were not evaluated statistically due to an insufficient number of cases. For comparisons of categorical data, chi-square or Fisher's exact tests were used, depending on whether an expected cell value was <5. Survival time was measured from the date of diagnosis. A P value of ≤.05 was considered significant in all tests.

The Kaplan-Meier procedure was used to compute nonparametric estimates of the survivor function and estimate survival days for cats with (SCC PROC LIFETEST, SAS version 9.1, SAS Institute Inc., Cary, NC). Cats (Nos. 11 and 12) were right censored if alive at the termination of the study.

Results

A total of 43 tumors were identified, representing 8 different tumor types. The age at diagnosis ranged from 1 to 16 years (mean, 10.4 ± 4.1 years). Significantly more tumors occurred in male cats than in females (27 and 16, respectively; P < .02). The right and left eyelids were nearly equally affected (19 right, 21 left). In 1 cat with AHC (No. 42), both the right and left eyelids were affected. In 4 cats the location (right/left) of the lesion was not specified. An equal number of tumors occurred on the upper and lower eyelids (17 upper, 16 lower). More tumors occurred on the medial side of the eyelid (9/15) than the lateral side (6/15). Six tumors occurred on the nictitans. In many cases, the lid location was not specified.

Squamous cell carcinomas

Eleven SCCs and 1 SCC in situ (No. 9) accounted for 28% of the tumors (Table 2). The average age of cats with SCCs was 12.4 ± 3.3 years. This was significantly (P < .05) higher than the average age (9.6 ± 4.2 years) of all other cats in the study. Based on evaluation of the adjacent non-neoplastic epithelium from 10 of the cats, 7 of the tumors arose in nonpigmented areas.

Table 2.

Eyelid squamous cell carcinomas: signalment, location, and outcome.∗

Case No.	Breed	Sex	Age (years)	Location	Pigmented Epithelium	Duration of Lesion Before Diagnosis	Complete Excision	Follow up
1	DSH	FS	5	Left lower	N	NA	N	NA
2	DSH	FS	14	Left lower	Y	6 months	N	NA
3	Domestic	F	14	Upper and lower	N	9 months	N	NA
4	DSH	MN	10	Right upper	NA	NA	N	Euthanatized due to tumor 1 month after diagnosis.
5	Mixed	MN	13	Nictitans	N	8 months	N	Euthanatized due to tumor 1 month after diagnosis.
6	DSH	F	8	Right lower	N	NA	N	Euthanatized due to tumor 2 months after diagnosis.
7	Domestic	F	15.5	Right medial canthus	N	NA	N	Euthanatized due to tumor 3 months after diagnosis.
8	DLH	MN	14	Right lower medial	N	3 months	N	Euthanatized due to tumor 4 months after diagnosis.
9	DSH	MN	16	Right upper	Y	Months	Y	Possible regrowth. Euthanatized 5 months after diagnosis.
10	Mixed	FS	14	Right lower medial	NA	>8 months	N	Treated with an iridium implant; 4 months later squamous cell carcinoma arose on nasal planum. Euthanatized due to tumor 17 months after initial diagnosis.
11	DSH	FS	11	Left lower medial	N	13 months	N	Right eye enucleated 10 months prior due to squamous cell carcinoma. Left eyelid previously biopsied and treated (cryotherapy, strontium) but still recurred. Cat still alive.
12	DSH	MN	14	Right upper	Y	Months	Y	Alive. DFI ≥ 23 months.
Average age at diagnosis			12.4

∗DSH = domestic shorthair; DLH = domestic longhair; F = female; FS = female spayed; MN = male neutered; NA = not available; Y = yes; N = no; DFI = disease-free interval.

The 9 cats for which follow-up information was available survived from 1 to >23 months after diagnosis. Of the 2 cats with completely excised tumors, 1 was still alive at the end of the study period (23 months later; No. 12); the other cat (No. 9) was euthanatized 5 months later for a possible recurrence. In 7 cases, the tumor resulted in euthanasia. The average survival time was 7.4 ± 2.5 months (Fig. 9). This value may be an underestimate, however, because it includes No. 12, which was still alive after 23 months. Had this cat been excluded, the average survival time would have been 5 ± 1.9 months.

Fig. 9.

Survival time of cats with eyelid squamous cell carcinomas (n  =  12).

Mast cell tumors

Eleven MCTs accounted for 26% of all tumors (Table 3). The average age at diagnosis was 6.5 years, which was significantly (P < .0001) lower than the average age (11.7 years) of cats with all other types of tumor. When available for evaluation, the overlying epithelium was pigmented in 8 of the 10 cats. One tumor (No. 5) was believed to be a recurrence of a tumor that had been incompletely excised 3 years previously; this cat was lost to follow up. Follow-up information was available for 8 of 11 cases. In all 8 cases for which follow-up information was available, the cats were still alive for at least 6 months to 8 years after diagnosis. No tumor recurred, despite incomplete excision in half the cases.

Table 3.

Eyelid mast cell tumors: signalment, location, and outcome.∗

Case No.	Breed	Sex	Age (years)	Location	Pigmented Epithelium	Duration of Lesion Before Diagnosis	Complete Excision	Follow Up
13	DSH	F	1	Left upper	Y	6 months	N	LTF
14	DSH	MN	9	Left upper	N	1.5 years	N	Treated (vinblastine, prednisone, CCNU). LTF 5 months after diagnosis.
15	DSH	F	8	Left upper	NA	6 months	N	Incompletely excised 3 years ago. At time of recurrence, there was a dermal MCT on shoulder. LTF
16	DSH	M	3.5	Right upper	Y	2 weeks	Y	Alive; DFI ≥ 8 months
17	DSH	M	11	Left lower lateral	Y	NA	Y	Alive; DFI ≥ 8 months
18	DSH	M	10	Left upper medial	N	1 month	N	Alive; DFI ≥ 10 months
19	DLH	M	4	Left upper	Y	NA	N	Alive; DFI ≥ 22 months
20	DSH	M	7.5	Left medial canthus	Y	1-2 months	N	Alive; DFI ≥ 19 months
21	DSH	M	7	Left upper medial	Y	2-3 months	Y	Alive; DFI ≥ 48 months
22	DSH	F	6	Left medial canthus	Y	1 year	N	Alive; DFI ≥ 73 months
23	Russian Blue	F	4	Right upper	Y	NA	Y	Alive; DFI ≥ 99 months
Average age at diagnosis			6.5

∗DSH = domestic shorthair; F = female; M = male; MN = male neutered; NA = not available; Y = yes; N = no; CCNU = lomustine; LTF = lost to follow up; DFI = disease-free interval.

Hemangiosarcomas

Six of the 43 tumors (14%) were HSAs (Table 4). Grossly, these were described as raised and red. Two HSAs occurred on the nictitans. The overlying epithelium lacked pigment in 5 of the 6 cats. The follow-up period for 3 of the cats ranged from 15 months to 5 years. All 3 of these cats had complete tumor excision, and none of the tumors recurred. The other 3 cats, whose tumors were not completely excised, were lost to follow up.

Table 4.

Eyelid hemangiosarcomas: signalment, location, and outcome.∗

Case No.	Breed	Sex	Age (years)	Location	Pigmented Epithelium	Duration of Lesion Before Diagnosis	Complete Excision	Follow Up
24	DSH	M	5	Right lateral commissure	N	NA	N	NA
25	DLH	FS	18	Nictitans	N	3 weeks	N	NA
26	DLH	MN	13	Right lower medial canthus	N	7 months	N	NA
27	DSH	FS	8	Nictitans	N	6 months	Y	Alive; DFI ≥ 15 months
28	DLH	MN	13	Left lower	Y	6 months	Y	Alive; DFI ≥ 16 months
29	DSH	MN	5	Right lateral canthus	N	NA	Y	DFI = 60 months, then LTF
Average age at diagnosis 10.3

∗DSH = domestic shorthair; DLH = domestic longhair; FS = female spayed; M = male; MN = male neutered; NA = not available; Y = yes; N = no; LTF = lost to follow up: DFI = disease-free interval.

Adenocarcinomas

ACAs accounted for 4 of 43 (9%) eyelid tumors (Table 5). One tumor (No. 33) had less distinct acini and a more basaloid cell morphology. The overlying epithelium in 3 of the 4 cats was pigmented. Postsurgical survival times were available for all 4 cats and ranged from 2 weeks to 12 months. In 1 cat, excision was complete; this cat lived 12 months before dying of unrelated causes. In the other 3 cases, tumor excision was not complete, and the cats were euthanatized or died because of the tumor. A necropsy was performed on 1 cat (No. 31); tumor metastases were present in the lung and regional lymph nodes.

Table 5.

Eyelid adenocarcinomas: signalment, location, and outcome.∗

Case No.	Breed	Sex	Age (years)	Location	Pigmented Epithelium	Duration of Lesion Before Diagnosis	Complete Excision	Follow Up
30	DSH	M	13	Nictitans	Y	1 year	N	Excisional biopsy 1 year earlier. Tumor recurred. Died 2 weeks after the second biopsy.
31	Persian	M	15	Right upper	Y	2 months	N	Euthanatized 2 weeks later.
32	DLH	F	12	Left lower	Y	3 months	N	Euthanatized because of tumor 2 months after diagnosis.
33	DLH	M	16	Right lower	N	2 months	Y	Died of other causes; DFI = 12 months.
Average age at diagnosis			14

∗DSH = domestic shorthair; DLH = domestic longhair; F = female; M = male; Y = yes; N = no; DFI = disease-free interval.

Peripheral nerve sheath tumors

Three PNSTs, each on an upper eyelid, accounted for 7% of all tumors (Table 6). All PNSTs had an Antoni A pattern, were negative for smooth muscle actin, and had varying degrees of S100 positivity. The overlying epithelium from 1 of the 3 cats was pigmented. Follow-up information was available for all 3 cases; the follow-up period ranged from 6 to 10 months. In all cases, excision was incomplete and the tumors recurred. One cat was euthanatized due to the tumor. Another cat died at home and was not necropsied. The third cat was still alive 6 months after the diagnosis.

Table 6.

Eyelid peripheral nerve sheath tumors: signalment, location, and outcome.∗

Case No.	Breed	Sex	Age (years)	Location	Pigmented Epithelium	Duration of Lesion Before Diagnosis	Complete Excision	Follow Up
34	DSH	M	15	Left upper	N	1-2 months	N	Tumor recurred; still alive 6 months later.
35	DSH	M	13	Right upper	Y	NA	N	Tumor recurred; died 7 months later.
36	DSH	M	10	Left upper	N	NA	N	Tumor recurred; euthanatized 10 months after initial diagnosis.
Average age at diagnosis			12.7

∗DSH = domestic shorthair; M = male; NA = not available; Y = yes; N = no.

Lymphoma

There were 3 lymphomas, representing 7% of the tumors (Table 7). The neoplastic cells did not stain with toluidine blue nor did they demonstrate immunoreactivity to antibodies for c-KIT (CD117). Two tumors were B-cell lymphomas as demonstrated by positive immunoreactivity for CD79a and negative immunoreactivity for CD3 and CD18. Positive immunoreactivity for CD3 and negative immunoreactivity for CD79a and CD18 confirmed that the third tumor was of T-cell origin. None of these tumors were definitively epitheliotropic; however, a Grenz zone was absent in the T-cell lymphoma (No. 38) and in 1 B-cell lymphoma (No. 39). The T cell tumor (No. 38) and one of the B cell tumors (No. 39) arose from the nictitans. The overlying epithelium was pigmented in 2 of the 3 cats. Follow-up information was available for 2 of the 3 cases. The follow-up period ranged from 5 to 6 weeks, and both cats were euthanatized. The reason for euthanasia was not specified for No. 38, whose tumor was incompletely excised. Euthanasia of No. 39 was attributed to tumor recurrence (despite apparently complete excision), unilateral renal enlargement, and deteriorating condition.

Table 7.

Eyelid lymphomas: signalment, location, and outcome.∗

Case No.	Breed	Sex	Age (years)	Location	Pigmented Epithelium	Type	Duration of Lesion Before Diagnosis	Complete Excision	Follow Up
37	DSH	M	14	Right upper	Y	B cell	3 weeks	N	Lymphoma near left scapula diagnosed 4 weeks previously. LTF
38	DSH	M	12	Right nictitans	N	T cell	3 days	N	Euthanatized 5 weeks later.
39	DSH	F	12	Left nictitans	Y	B cell	7-10 days	Y	Euthanatized 6 weeks after diagnosis due to anorexia, weight loss, enlarged kidney, and tumor recurrence.
Average age at diagnosis			12.7

∗DSH = domestic shorthair; F = female; M = male; Y = yes; N = no; LTF = lost to follow up.

Apocrine hidrocystomas

Three AHCs accounted for 7% of the eyelid tumors (Table 8). Grossly, they were described as raised and often darkly pigmented masses. In 2 of the 3 cats, the tumors arose beneath pigmented epithelium. The follow-up period for all 3 cats ranged from 2 to 5 years. All tumors were completely excised. Neither of the 2 domestic shorthair cats experienced recurrence, and both died or were euthanatized for reasons other than the tumor. The Persian cat (No. 42) had a 3-year history of the tumors; 5 years later, additional tumors were present.

Table 8.

Eyelid apocrine hidrocystomas: signalment, location, and outcome.∗

Case No.	Breed	Sex	Age (years)	Location	Pigmented Epithelium	Duration of Lesion Before Diagnosis	Complete Excision	Follow Up
40	DSH	MN	9	Lower eyelid	N	1 month	Y	Similar tumor on ear pinna was removed at the time of diagnosis. Died of other causes. DFI = 24 months.
41	DSH	FS	9	Lower eyelid	Y	NA	Y	Euthanatized for other reasons. DFI = 48 months.
42	Persian	MN	5	Left medial canthus	Y	3 years	Y	Had multiple tumors removed at the time of diagnosis; some tumors recurred after 60 months. LTF
Average age at diagnosis			7.7

∗DSH = domestic shorthair; FS = female spayed; MN = male neutered; NA = not available; Y = yes; N = no; LTF = lost to follow up; DFI = disease-free interval.

Hemangioma

A single hemangioma accounted for 2% of all tumors. This tumor (No. 43) arose on the lower, right eyelid of an 11-year-old, neutered male, domestic shorthair cat. The tumor was described as raised, firm, and nodular. The overlying epithelium was pigmented. The tumor reportedly had been present approximately 6 months before its complete excision. There was no tumor recurrence during the 16-month follow-up period. At the time of this report, the cat was clinically healthy.

Discussion

The prevalence of eyelid tumors in this study varies considerably from that in the last comprehensive study ¹⁷ (Table 9). Notably, MCTs were more common in the current study, and the number of SCCs declined. The decreased prevalence of SCC may be due in part to an increased awareness of SCCs by general practitioners and a failure to biopsy these lesions in all cats. The reason for the increased numbers of MCTs is unclear.

Table 9.

Comparison of the prevalence of eyelid tumors in this study with that in previous studies.¹⁷

Tumor Type	UTCVM∗	VMDP	PCOP
Hemangioma	1/43 (2%)	1/86 (1%)	-
Apocrine hidrocystoma	3/43 (7%)	-	1/36 (2.8%)
Lymphoma	3/43 (7%)	-	4/36 (11.1%)
Malignant peripheral nerve sheath tumor	3/43 (7%)	-	-
Adenocarcinomas	4/43 (9%)	3/86 (3.5%)	-
Hemangiosarcoma	6/43 (14%)	2/86 (2.3%)	-
Mast cell tumor	11/43 (26%)	-	4/36 (11.1%)
Squamous cell carcinoma	12/43 (28%)	56/86 (65%)	13/36 (36.1%)

∗UTCVM = University of Tennessee College of Veterinary Medicine; VMDP = Veterinary Medical Data Program; PCOP = Purdue Comparative Oncology Program.

The average age at diagnosis was 10.4 years. This is consistent with the previous report in which eyelid tumors were more frequent in cats older than 10 years. ¹⁷ Significantly more males than females were affected (P < .02). Approximately equal numbers of tumors arose on the right versus left eyelids, upper versus lower lid, and medial versus lateral sides of the lid.

Six tumors arose from the nictitans, including 2 HSAs, 2 lymphomas, 1 SCC, and 1 ACA. A previous report of tumors in the nictitans of cats found 2 SCCs, 2 ACAs, and 1 lymphoma. Both ACAs in that study recurred within 3 to 7 months and, in 1 case, resulted in euthanasia of the cat. ²² In the current study, the single ACA arising from the nictitans (No. 30) recurred a year after the initial diagnosis and resulted in the death of the cat 2 weeks later.

Neoplasms of squamous epithelium were the most common eyelid tumor in this study and accounted for 28% (12/43) of the tumors. This is less than in a previous study, which reported a prevalence of 65% (56/86) and 36% (13/36) in the VMDP and PCOP studies, respectively. ¹⁷ Although cats are predisposed to SCCs of the eyelids, lesions may simultaneously occur on the ear pinnae, nasal planum, or lips. ⁸ Although the coat color of the cats was not available in most cases, the histologic presence of pigment in the surrounding non-neoplastic epithelium was used as a crude indicator of coat color. Seven of 10 SCCs arose from nonpigmented epithelium; this is consistent with the proposed ultraviolet-mediated pathogenesis of these tumors and the predisposition of white cats to SCCs. ⁸ Only 2 of the cats in the current study had a history of SCC or concurrent SCC in other areas.

Cats with SCCs were significantly older than cats with all other tumor types combined (12.4 years versus 9.6 years; P < .05). This is consistent with a report stating that the average age of cats with cutaneous SCCs was 11 years. ⁸ Similarly, the study describing eyelid tumors at the PCOP and VMDP found that the risk of developing SCCs significantly increased in cats older than 10 years. ¹⁷

Metastases to regional lymph nodes are reported to occur late in the course of SCC. ²⁴ None of the euthanatized cats in this study were submitted for necropsy; hence, the metastatic potential of these SCCs could not be evaluated. In 1 report, 13 of 61 (21%) of affected cats with nasal or pinnal SCCs died or were euthanatized due to the tumor. ⁸ In the current study, 7 of 9 (78%) cats were euthanatized due to their tumors. SCCs were significantly more likely to recur than MCTs (P < .006) and had an average survival time of 7.4 months. This survival time may be an underestimate, because the longest surviving cat was still alive at the end of the study period.

Eleven MCTs were identified in this study, representing 26% of all the eyelid tumors. This is a much higher prevalence than in the previous report in which 4 of 36 (11.1%) eyelid tumors from the PCOP and 3 of 86 (3.5%) eyelid tumors from the VMDP were MCTs. ¹⁷ The reason for this apparent increase in prevalence is unclear.

Feline cutaneous MCTs typically occur in older cats and are most commonly present on the head and trunk. ^4,10 Of those arising on the head, the most common sites are the temples, pinnae, periorbital areas, ⁴ and haired skin of the eyelid. ^23,25 In cats, cutaneous MCTs tend to have a benign clinical course, even when they are histologically pleomorphic. ¹⁵ Mitotic index has been shown to be the most effective prognosticator of the behavior of feline cutaneous MCTs. ^15,16

The average age at diagnosis of cats with eyelid MCTs in this study was 6.5 years. This was significantly (P < .001) lower than the 11.7 years for cats with all the other types of eyelid tumors in this study. This age is also younger than that reported for feline cutaneous MCTs. One study reported an average age of onset for feline cutaneous MCT of 9 to 11 years, with MCT rarely occurring in cats younger than 4 years. ¹⁰ Cutaneous MCTs in younger cats are more likely to be atypical or poorly granulated forms. ¹⁰ These atypical/poorly granulated forms also tend to occur in Siamese cats, a breed known to develop multiple cutaneous MCTs. ¹⁰ None of the tumors in the current study were pleomorphic or poorly differentiated. There was no breed predisposition noted in the current study, because only 1 tumor arose in a purebred (Russian Blue) cat. As with MCTs in skin, eyelid MCTs in this study had benign behavior and did not recur, even after incomplete excision.

In a previous report, 6 periocular PNSTs were identified over a 26-year period (0.15% prevalence). ¹³ In the current report, 3 of 43 tumors (7%) were PNSTs. All occurred on the upper eyelid, which is consistent with the previous study in which the upper eyelid was more commonly affected. ¹³ None of these tumors was completely excised, and, as in the previous study, the tumors were locally invasive with a 100% recurrence rate.

In a previous review, 2 of 86 tumors (2.3%) were HSAs. ¹⁷ In the current report, 14% (6/43) tumors were HSAs. Another group reported 6 hemangiomas and 2 HSAs arising from the conjunctiva or third eyelid of cats; both HSAs recurred after excision, but re-excision of the tumors was successful. ²⁰ Additionally, single cases of HSAs arising from the lower eyelid ¹² and third eyelid ¹⁹ did not recur after complete excision. In this study, all HSAs, for which follow up was available, were completely excised and none recurred. Despite their histologic features of malignancy, HSAs arising from the eyelids of cats appear to have a favorable outcome when completely excised. Because cats with incompletely excised HSAs were all lost to follow up, the clinical behavior of incompletely excised HSAs cannot be inferred from this study.

Five of the 7 vascular tumors (hemangioma and HSAs) in this study arose in males. Previous reports of vascular tumors in the conjunctiva ²⁰ and skin ¹⁸ demonstrated an increased prevalence in male cats. Despite the trend observed in this study, this difference was not statistically significant.

As is the case with vascular tumors arising in the skin, recent reports have suggested that conjunctival hemangiomas and HSAs in both dogs ⁶ and cats ²⁰ may be due to exposure to ultraviolet radiation. In support of this, one study found that 87% of feline conjunctival hemangiomas and HSAs arose in nonpigmented areas. ²⁰ Similarly, in the current study, 5 of 6 HSAs arose beneath an unpigmented epithelium.

Adenocarcinomas represented 4 of 43 (9%) of all eyelid tumors in this study, compared with 3 of 86 (3.5%) in a previous study. ¹⁷ The incompletely excised ACAs in this study invariably resulted in the death or euthanasia of the cat.

Three cases of eyelid lymphoma were identified in this study. The previous review found 4 cases of eyelid lymphoma (4/36; 11.1%). In addition to a report of lymphoma arising from the third eyelid of a cat, ²¹ there is a single case report of conjunctival Hodgkins-like lymphoma in the upper eyelid of a cat. ¹⁴ After surgical excision, chemotherapy, and radiation therapy, the tumor did not recur. ¹⁴ In the current study, there were 1 T-cell lymphoma and 2 B-cell lymphomas. Epitheliotropism was not obvious in any tumor. Both cats, for which follow up information was available, were euthanatized within 6 weeks of diagnosis. In neither case was the reason for euthanasia explicitly stated, nor was a necropsy performed; however, based on these results, eyelid lymphoma in cats is uncommon but may carry a poor prognosis.

In this study, there were 3 AHCs (7%) compared with a single cystadenoma (1/36; 2.8%) in the previous report. ¹⁷ Hidrocystomas also occur in humans ^1,5 and are thought to arise from eccrine or apocrine sweat glands. ² Whether these lesions arise secondary to inflammation or duct obstruction or whether they are truly benign neoplasms is debatable. These lesions have also been referred to as cystic apocrine hyperplasia. ²⁴ The most recent report of these lesions in cats concluded that they were in fact adenomas, based on the presence of papillary luminal projections and proliferative activity, indicated by positive Ki67 staining, and the absence of inflammation. ³ In that study, 5 of the 6 cats were Persians, which suggested a possible breed predisposition. ^3,5 In the current report, 1 of 3 cats was Persian. The Persian had several tumors that began at 2 years of age. The increased number of tumors in the Persian cat in this study and the young age at onset of the lesions support a hypothesis of a breed predisposition to AHCs.

Because these tumors were described clinically as darkly pigmented, it has been suggested that melanomas should be considered as a differential diagnosis; ³ however, the results of this and other studies ¹⁷ indicate that eyelid melanocytic neoplasms are uncommon in cats. Based on the findings in this study, cats may have multiple AHCs and additional tumors may arise after excision.

Both SCCs and HSAs are proposed to have an ultraviolet-mediated pathogenesis. ^8,20 In the current report, 70% and 83% of these tumors, respectively, arose from nonpigmented areas. In contrast, 8 of 10 MCTs arose beneath pigmented epithelium (MCT development has not been correlated with ultraviolet exposure or the presence of pigment.) When compared with all other tumors combined, SCCs and HSAs were significantly (P  =  .0089) more likely to occur in nonpigmented areas. This finding supports their ultraviolet-mediated pathogenesis, because melanin pigment is thought to have a protective effect in ultraviolet-mediated carcinogenesis.

Feline eyelid tumors were significantly more common in male cats. Older cats are primarily affected, but younger cats are more likely to have MCTs than SCCs. SCCs and HSAs are significantly more likely to arise in nonpigmented areas. Tumors that are histologically malignant (HSA, lymphoma, ACA, SCC) or potentially malignant (MCT, PNST) accounted for 39 of 43 (91%) of tumors identified in this study. This is consistent with a previous report stating that eyelid tumors in cats are more likely to be malignant than benign. ²⁶ In this study, the hemangioma and HSAs did not recur after complete excision. Most MCTs did not recur, regardless of whether or not they were completely excised. Cats with AHCs tended to have multiple lesions, but the clinical course was benign after complete excision. In contrast, lymphomas, ACAs, SCCs, and PNSTs frequently resulted in death or euthanasia of the cat. These more aggressive tumor types accounted for 22 of 43 (51%) of the tumors in this study. Therefore, eyelid tumors in cats should be evaluated histologically to confirm the diagnosis; complete excision of all tumors is recommended.