Melanoma

From Felipedia

Melanoma are a rare and highly metastatic skin disease of cats.

A diagnosis of melanoma in domestic animals typically carries a grave prognosis. It is generally detected at a late stage when excision is rarely curative and metastasis is often already detectable within regional lymph nodes (LN). In dogs it is a common diagnosis, accounting for 7% of all malignant tumors. It is the most common malignant neoplasm of the oral cavity and the second most frequent subungual neoplasm. In cats it is not as common but carries the same poor prognosis, accounting for less than 1% of all feline oral neoplasms and approximately 0.5% of feline skin tumors. The ocular and cutaneous forms are generally more common than melanoma of the oral cavity. The most common cutaneous sites are head, tail, distal extremities, and lumbar area. The prognosis is generally poor because of recurrence and regional metastasis in up to half the cases. In a study of 23 cats with non-ocular melanocytic tumours, approximately half were malignant^[1].

A number of variants are known to occur:

• Melanocytic hyperplasia - a non-neoplastic proliferation of melanocytes within the epidermis, primarily the basal cell layer.
• Melanocytoma - a benign tumor arising from melanocytes. These cells may be in the epidermis, dermis, or adnexa, primarily the hair follicles
• Malignant melanoma - as in humans, highly metastatic and incurable
• Ocular melanoma

Criteria of malignancy

Melanoma may be one of the few neoplasms in animals for which location is an important prognostic indicator in its own right. Melanocytic neoplasias involving the oral cavity, subungual region, and mucocutaneous junctions are considered malignant, regardless of any other single feature. Interestingly, this also seems to be the case with oral melanoma in humans. There is no evident relationship between histologic characteristics, including mitotic index and pigmentation, and survival rate. The prognosis for human melanoma at mucocutaneous junctions and mucosal surfaces is also grave.

In animal skin and eye melanocytic neoplasms, the most reliable histologic feature for distinguishing malignant from benign is the mitotic index. In the World Health Organization's Histologic Classification of Epithelial and Melanocytic Skin Tumours of Domestic Animals, three or more mitotic figures per 10 high-power fields indicate malignancy. The identification of mitotic figures in sections must be undertaken with great care to ensure that only true mitoses are counted. Bleached sections should be used for pigmented neoplasms to avoid confusing mitoses with pyknotic nuclei and the small hyperchromatic nuclei of the spindle cells of the supporting stroma. The mitotic count will vary from area to area within the neoplasm, and the count should be performed in areas with the greatest concentration of mitotic figures. Where metastases have been confirmed, the mitotic rate in intraocular melanoma is usually greater than four per 10 high-power fields; less than two mitotic figures per 10 high-power fields is consistent with a melanocytoma. In cats the value of mitotic activity to prediction of outcome is less clear, particularly with regard to ocular melanoma. Some authors, however, report a link between high mitotic index and increased risk of metastatic disease^[2].

Neoplastic cell morphology is also a useful discriminating feature. Cytologic features of malignancy include the presence of a large nucleus, variation in nuclear size and shape, hyperchromasia, abnormal chromatin clumping, one or more nucleoli, and atypical mitotic figures. Additional features favouring malignancy are the presence of neoplastic cells, individually or in nests, within the upper layers of the epidermis, although ulceration and necrosis may prevent evaluation of this feature. The presence or absence of junctional activity is not specific to melanoma and often occurs in melanocytomas.

The gold standard is, of course, lymphatic or vascular invasion. Pathologists are often asked to evaluate mandibular LN for evidence of metastatic disease in dogs with oral melanoma. These nodes may be pigmented because of metastatic melanoma or the accumulation of melanophages within the medullary sinuses. The latter occurs when melanophages are transported to the regional LN after inflammation of the oral mucosa, particularly in dogs with pigmented oral epithelium. The cytologic features described above, when used to evaluate bleached sections, usually suffice to distinguish melanophages from melanoma.

In cats the following histologic features have been proposed to be of significance in identifying cutaneous melanoma: nuclear atypia (neoplasms with more extensive atypia are more likely to be malignant), mitotic activity (a trend of greater malignancy with increased numbers of mitoses), and tumour cell type (with epithelioid more likely to be malignant). The relationship with cell type is not firmly established in the literature, in that two studies indicate a less precise relationship between cell type and malignant potential. Furthermore, histologic determination of malignancy does not consistently correlate with clinical behavior.

The Callender system has been used historically to determine malignant potential in canine anterior uveal melanocytic neoplasms. This system was designed to predict the behaviour of ocular melanoma in humans, which led to criticisms of its use in the veterinary literature for two main reasons. Firstly, it was designed for malignant tumours in humans, and most canine intraocular melanocytic tumours are benign. Secondly, the morphology of the canine neoplastic melanocytes does not match the cell descriptions in the Callender classification. Several authors favour a simpler system, dividing intraocular melanocytic tumours into benign and malignant using well-recognized cytologic features of malignancy as a basis for this classification. Despite these controversies, some useful morphologic features can be used to ascribe malignancy, especially with canine anterior uveal melanoma. A greater risk of metastasis tends to occur in mixed and epithelioid cell types. Melanocytic neoplasms containing narrow spindle-shaped cells with small nuclei and no mitotic figures, or plump cells with large amounts of melanin and small round nuclei, are more likely to be benign. This is in contrast to the cat where cellular pleomorphism, degree of pigmentation, nuclear to cytoplasmic ratio, and number of nucleoli bear no relationship to the metastatic behaviour of uveal melanocytic tumors[1].

Tumor size, degree of pigmentation, intensity of proliferating cell nuclear antigen (PCNA)–staining, the presence of necrosis, ulceration, or inflammation, and p53 expression are of limited prognostic value in animals. Depth of tumour invasion into the skin is an important criterion in human medicine but is felt to be too impractical in animals. More sophisticated techniques, such as flow cytometry to detect chromosomal anomalies, offer no real advantage over histology in predicting tumor behavior. The value of determination of a proliferative index using MIB-1, a monoclonal antibody to Ki-67, which identifies proliferating cells, has been investigated. One study examined 27 canine and feline melanocytic tumours, eight benign and 19 malignant, to correlate Ki-67 and PCNA activity with 6-month survival. Ki-67 activity correlated very well with poor survival but so did the presence of invasive growth and classification based on cytologic criteria. PCNA was significantly higher in malignant neoplasms, but levels did not demonstrate a strong correlation with poor survival. In a separate study, MIB-1 was applied to 68 cutaneous canine melanocytic neoplasms to determine if the level of immunoreactivity would correlate with 2-year survival. Eighteen of 68 tumours were classed as malignant histologically. The predictive value of the Ki-67 index for 2-year survival was 97%, which was only slightly higher than the predictive value associated with histologic evaluation (91%)^[3].

Differential diagnosis

Other tumours can look clinically very similar to melanoma, particularly those arising from the skin. These include melanocytoma as well as pigmented lesions of the epidermis and adnexa, e.g., feline basal cell tumour and carcinomas, trichoblastoma, trichoepithelioma, pilomatricoma, sebaceous adenoma, and apocrine neoplasms. Ceruminous cysts in the feline may be mistaken for multicentric melanoma of the pinna. Melanocytic hyperplasia (lentigo simplex) on the lips, eyelids, nose, and gingiva of orange, cream, and silver cats appear as pigmented macules at these sites. Hyperpigmented macular lesions clinically resembling melanoma may occur in the skin in dogs, primarily the abdomen and nipple. Similarly, epidermal hamartomas (pigmented epidermal nevi, canine seborrheic keratosis), and dermal hemangioma and hemangiosarcoma can appear as pigmented cutaneous tumors.

Non-melanocytic neoplasms that commonly arise in the oral cavity and eye seldom present clinically as pigmented tumours; therefore, pigmentation is not a distraction.

References

1. ↑ VetPathology.org
2. ↑ VetPathology.org
3. ↑ S. H. Smith, M. H. Goldschmidt and P. M. McManus (2002) A Comparative Review of Melanocytic Neoplasms. Vet Pathol 39:651-678