Purpose: Primary orbital melanomas are rare tumors with a poorly defin
ed biologic course. Most recorded experiences concern single case repo
rts. The authors evaluated the applicability of several of the histopa
thologic prognostic indicators used for uveal melanomas to a series of
primary orbital melanomas with known clinical follow-up. Methods: Twe
nty-one primary orbital melanomas, each with at least al-year followup
after diagnosis, were evaluated for (1) modified Callender cell type,
(2) mitotic count per 40 high-power fields, (3) lymphocyte count (les
s than versus greater than 100/20 high-power fields), (4) blue nevus c
omponent, and (5) largest tumor diameter. Results: All patients for wh
om race was recorded were white. The mean age at diagnosis was 42 year
s (range, 15-84 years). There was an associated blue nevus in 19 patie
nts (90%), and in 10 patients (47.5%) there was some form of congenita
l melanosis. With a mean follow-up period of 4.5 years (range, 1-13 ye
ars), mortality from metastatic tumor occurred in 8 (38%) of 21 patien
ts. Of these eight patients, there were liver metastases in seven (88%
) and brain metastases in one (12%). Indicators of poor prognosis were
tumors of mixed cell type with high mitotic count and greater patient
age with underlying congenital melanosis. Conclusion: Most primary or
bital melanomas occur in white patients and are associated with blue n
evi. These tumors are similar to uveal melanomas with respect to progn
ostic indicators and pattern of metastasis.