The prognostic value of tumor blood vessel morphology in primary uveal melanoma.
BACKGROUND:It is possible to identify at least nine vascular patterns in melanomas of the ciliary body and choroid from histologic sections. An association between the presence of at least one closed vascular loop and death from metastases was shown in a matched-pair, case-control study of 40 patients whose eyes were removed for ciliary body or choroidal melanomas. METHODS:Two independent observers who were masked to the follow-up of patients examined histologic preparations of 234 eyes removed for ciliary body or choroidal melanomas for the presence of each of the tumor vascular patterns. Statistical analyses included tests for interobserver reliability, Kaplan-Meier survival curves, and the fitting of Cox regression models. RESULTS:The detection of each of the nine vascular patterns is highly reproducible. The Cox model indicates that the presence of vascular networks, defined as at least three back-to-back closed vascular loops, is the feature most strongly associated with death from metastatic melanoma. Other significant factors in the Cox model include (in descending order of importance) largest tumor dimension, mitoses, the parallel with cross-linking vascular pattern, age, the presence of tumor-infiltrating lymphocytes, and male gender. CONCLUSIONS:The presence of vascular networks provides the most significant association with death from metastatic melanoma of all variables tested. The presence of this pattern should be recorded on pathology reports. If it becomes possible to detect this vascular pattern clinically using a noninvasive imaging technique, then ophthalmologists may be able to determine the likely biologic behavior of a melanoma before resorting to the removal of tissue.
Folberg, R; Rummelt, V; Parys-Van Ginderdeuren, R; Hwang, T; Woolson, RF; Pe'er, J; Gruman, LM
Volume / Issue
Start / End Page
Pubmed Central ID
Electronic International Standard Serial Number (EISSN)
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)