We retrospectively reviewed the microscopic findings in 32 histologically c
onfirmed radial scars in 31 women diagnosed in our unit during 1994-1998. T
he median age at diagnosis was 53 years (range 47-63 years). Thirty-one (97
%) of 32 lesions presented as screen detected mammographic abnormalities (2
8 stellate lesions, 2 microcalcifications and only 1 architectural distorti
on). One lesion presented as a palpable breast mass. Stereotactic or ultras
ound-guided fine nee die aspiration cytology (FNAC) was performed in 28 cas
es. Cytological analysis of FNAs revealed malignant cells (CS) in 8 (29%) c
ases, highly suspicious cells (C4) in 3 (11%)cases, atypical benign cells (
C3) in 7 (25%) cases and benign epithelial cells (C2) in 10 (35%) cases.
All non-palpable lesions were surgically excised following wire localizatio
n. Histological examination of the breast specimens (mean weight = 16 g) de
monstrated, in addition to a radial scar, 6 invasive carcinomas (2 infiltra
ting ductal, 2 tubular, 1 mixed ductal/lobular and 1 secretory carcinoma) a
nd 4 ductal carcinoma in situ lesions (2 high grade, 1 high grade with micr
o-invasion and I low grade) arising in the radial scar. Of the remaining ca
ses the radial scar was associated with atypical epithelial hyperplasia in
2 cases and regular epithelial hyperplasia in 17 cases (4 florid and 13 mil
d to moderate). In the 10 cases associated with malignancy, 9 had FNAC and
was reported as malignant (C5) in 6 (67%) cases, highly suspicious (C4) in
2 (22%) cases and atypical (C3) in 1 (11%). False positive FNAC (C5) occurr
ed in two patients, one of whom presented with pleomorphic microcalcificati
ons suggestive of ductal carcinoma in situ. This patient was treated with a
wire guided segmental mastectomy. All invasive tumours were less than 20 m
m in size (T1) and of these 4 were grade I and 2 were grade II. Axillary di
ssection was performed in 4 patients none of whom had axillary node metasta
ses.
Our study demonstrates a significant incidence of malignancy associated wit
h radial scars (31%) suggesting that radial scars may be premalignant lesio
ns. This is supported by detecting various stages of mammary carcinogenesis
(atypical epithelial hyperplasia, ductal carcinoma in situ, and early inva
sive malignancy) in these lesions. Fine needle aspiration cytology seems to
be unreliable in the diagnosis of radial scar associated malignancy (67% s
ensitivity and 91% specificity). Stellate lesions, therefore, should be exc
ised to obtain an histological diagnosis regardless of cytological findings
. Further studies examining the biology of radial scars are required. (C) 1
999 Harcourt Publishers Ltd.