FNA biopsy of 51 orbital masses is critically reviewed. Aspiration was
performed with a 23 G needle inserted by an ophthalmologist; the smea
rs were prepared by a cytologist. Forty-two cases (83%) were correctly
diagnosed as benign or malignant either with (68%) or without (15%) c
orrect specification of the histology, There were two false-negative a
nd seven inadequate cases. Immunocytochemical stains were performed in
five cases using the following antibodies: L26 (Pan B), UCHL1 (Pan T)
, kappa and lambda immunoglobulin light chains (three cases) in order
to distinguish inflammatory pseudotumours from low-grade non-Hodgkin's
lymphomas, In two cases we used CAM 5.2 (a monoclonal cytokeratin coc
ktail) and vimentin to ascertain the epithelial origin of two metastat
ic tumours, In five other cases cytospins were not adequately cellular
for immunocytochemistry. Insufficient material and one false-negative
sample were obtained from very fibrotic lesions or from posteriorly l
ocated lesions. The results are discussed and compared with other seri
es reported in the literature. Orbital FNA biopsy may be considered a
useful tool in the diagnostic approach to orbital masses in which the
relatively high number of inadequate aspirations is offset by a low co
st-benefit ratio.