No abstract

Abstract. Primary ocular melanocytic neoplasms from 9 1 dogs were divided into two groups by histologic criteria. Seventy-five were benign and composed of spindle-shaped and large polyhedral melanocytes similar to those of human ocular melanocytomas. Fifty-nine of these originated in the uvea where most resulted in uveitis, glaucoma, or hyphema prior to enucleation. None metastasized. Nineteen melanocytomas were limbal tumors. None metastasized, but three of nine incompletely excised tumors were found within the anterior chamber 2 to 3 years after the initial removal. Sixteen uveal melanocytic neoplasms were histologically malignant. Three had confirmed metastases, all within 3 months of enucleation. Cell type or pattern of growth within the globe were not predictive of biologic behavior. Our data suggest that the mitotic index is the best criterion for histologic identification of ocular melanomas with high metastatic potential. We propose that the classification of primary ocular melanomas be simplified to include only two categories: melanocytoma (benign) and melanoma (potentially malignant). Further behavioral data may justify a grading scheme for melanomas based upon mitotic index.One rationale for the detailed histologic classification of neoplasms is the belief that at least some aspects of tumor morphology are predictive of biologic behavior or response to therapy. Debate stems from the degree of refinement necessary for prognostic accuracy, debate stimulated in part by the persistence of cumbersome, descriptive classifications of unproven prognostic merit. For those few animal tumors that have been adequately studied, behavior has been better predicted by cytologic indices of malignancy (mitotic index, nuclear atypia, nuclear : cytoplasmic ratio) than by classification based upon cell type and architectural arrangement. [1][2][3]5,18 The same cytologic indices will not likely be equally valid for every type of neoplasm. Some tumors habitually behave in ways not expected from their cytologic characteristics, and the same tumor occurring in different sites may show marked differences in behavior, especially canine melanocytic tumors. The histologic classification of skin melanomas into groups defined by cell type or architectural arrangement is of little prognostic value. Mitotic index is a more accurate predictor of behavior. In contrast, melanomas originating in the mouth behave in an aggressive manner not correlated to any identified histologic or cytologic parameter. Primary ocular melanomas have been subjected to more debate but less consensus than have melanomas of any other site. Most authors have used an adaptation of Callender's classification for human uveal melanomas, a scheme based upon cell type that is very predictive of biologic behavior in p e~p l e .~J~-'~ Whether or not the Callender scheme is appropriate and accurate for canine tumors has not been established because so few canine tumors have had confirmed m e t a s t a~e s .~~~~, '~~~~ Furthermore, many canine ocular melanomas are hist...

Bovine pericardium graft offers a promising option for surgical reconstruction of the cornea following keratectomy for the management of corneal ulcers and sequestra.

The objectives of this study were to evaluate morphologic, histochemical, and immunohistochemical characteristics of well-differentiated and anaplastic intraocular neoplasms of cats, and to develop a diagnostic algorithm for, and investigate the association of ruptured lenses with these neoplasms. Seventy-five feline globes with intraocular neoplasms were stained with hematoxylin and eosin and examined by light microscopy. Morphologic diagnoses included 33 intraocular sarcomas, 17 diffuse iris melanomas, 15 lymphosarcomas, three ciliary adenomas, one metastatic carcinoma, and six undifferentiated intraocular neoplasms. Sections of these globes were then stained with periodic acid Schiff (PAS), and immunohistochemical (IHC) labels for various cellular markers. Histochemical staining and IHC labeling confirmed cellular differentiation in 73/75 neoplasms but was discordant with morphologic diagnoses in 8/75. These included four neoplasms morphologically diagnosed as lymphosarcomas but which expressed differentiation antigens consistent with melanoma (n = 3) or ciliary adenocarcinoma (n = 1), and four tumors morphologically diagnosed as intraocular sarcomas that expressed differentiation antigens for melanoma (n = 2), metastatic carcinoma (n = 1), or remained undifferentiated (n = 1). Immunohistochemical labeling suggested a diagnosis in 5/6 morphologically undifferentiated neoplasms including one intraocular sarcoma, two diffuse iridal melanomas, and two ciliary adenocarcinomas. Based upon morphologic, histochemical, and IHC characterization, ruptured lens capsules were detected in 28/30 intraocular sarcomas, 3/24 diffuse iris melanomas and 1/11 lymphosarcomas, but not in ciliary epithelial neoplasms, metastatic carcinomas, or undifferentiated intraocular neoplasms. An algorithm is provided that facilitates stain and IHC label selection for differentiating anaplastic intraocular feline neoplasms.

No abstract