RESUMO
The incidence of malignancy in excision biopsies performed for atypical ductal hyperplasia (ADH) diagnosed on needle biopsies has decreased since the advent of larger tissue sampling and improved accuracy using vacuum-assisted Mammotome biopsy. We undertook a retrospective study to identify predictive factors for understaging of ADH diagnosed on 11-gauge Mammotome biopsy, to determine whether it was possible to avoid surgical excision in women where mammographically visible calcifications had been completely removed. Sixty-one biopsy diagnosed ADH lesions were correlated with surgical excision findings that revealed DCIS in 14 (23%). The mammographic and biopsy features were statistically analyzed using Fisher's exact test. There was no association between morphology, extent of calcifications, number of cores sampled with underestimation of malignancy (P=0.503, 0.709, 0.551 respectively). In the absence of residual calcifications, the frequency of underestimation of carcinoma still occurred in 17%.
