Aims: The purpose of this study was to test different malignant non-melanoc
ytic tumours with the commercially available antibody Melan-A to examine it
s diagnostic specificity and to compare the S100, Melan-A and HMB-45 reacti
vity in various melanocytic lesions.
Methods and results: Seventy-three benign and malignant melanocytic lesions
and 31 cases of non-melanocytic tumours, sarcomas, carcinomas and carcinoi
ds, were selected. Immunohistochemical staining of paraffin sections, follo
wing a high temperature antigen unmasking technique, was performed. Melan-A
stains junctional and dermal melanocytes in all benign melanocytic lesions
with the exception of neuro-naevoid areas, The epithelioid and the spindle
cells in malignant melanomas did not show considerable difference in their
Melan-A reactivity. The predominantly spindle cell type mucosal melanomas
contained more Melan-A-positive cells than HMB-45-positive cells and simila
r results were observed in metastatic malignant melanomas. In desmoplastic
melanomas the positivity of Meran-A was not consistent. None of the sarcoma
s, carcinomas and carcinoids expressed Melan-A. Almost all soft tissue tumo
urs, except for two malignant gastrointestinal stromal tumours, were unreac
tive for HMB-45. These two cases did not react with hilelan-A antibody.
Conclusions: Melan-A is a useful additional marker to differentiate non-mel
anocytic tumours from primary or metastatic melanoma, In melanocytic lesion
s the Melan-A staining pattern is similar to S100, but seems to be more spe
cific. In desmoplastic melanomas, however, the variable Melan-A staining fu
rther necessitated detailed histological examination and the use of the S10
0 reaction.