INTRODUCTION: Flat Epithelial Atypia (FEA), Atypical Ductal Hyperplasia (ADH), and Atypical Lobular Hyperplasia (ALH) are well known non-obligate precursors of breast malignancy. Management of FEA, ADH and ALH after image-guided needle biopsy (NB) still remains controversial. The aim of this study was to evaluate the rate of malignancy underestimation associated to FEA, ADH or ALH diagnosis by image-guided NB, either core needle biopsies (CNB) or vacuum assisted breast biopsy (VABB). As a secondary objective, we attempted to identify clinical characteristics and imaging features associated with malignancy.
MATERIAL AND METHOD: We retrospectively reviewed all consecutive image-guided NB performed between January 2006 and April 2016 that resulted in a diagnosis of FEA, ADH or ALH which were subsequently submitted to surgical biopsy. All pathological reports were available. We compared the pathology specimens obtained from biopsy to those obtained from surgery. The rate of underestimation was calculated by means of specific mathematical equations.
RESULTS: Among 4715 patients who underwent image guided NB, 174 (3.7%) resulted in diagnosis of FEA, ADH and ALH. 49 patients were excluded because of loss in follow up.The diagnosis in the remaining 126 cases, were 45 (36%) FEA, 66 (51%) ADH, 9 (7%) ALH and 8 (6%) combined diagnoses. Of these, 16 (13%) were ultrasound-guided CNB procedures and 110 (87%) stereotactic VABBs. In thirty-six cases (28,3%) the diagnosis was upgraded to malignancy. For CNB specimens, we found underestimation in 10 of 16 procedures (62.5%) and at VABB there were 26 cases of underestimation out of a total of 110 biopsies (23%). Univariate analysis indicated that use of CNB (p=0.001), residual calcifications (p=0.008), and presence of ALH in NB specimens (p=0.08) were independent predictors of underestimation. The presence of a mass on ultrasound was more likely to be associated with malignancy, as 5 out of 10 lumps (50%) resulted in lesion upgrade. Neither age at diagnosis nor hormonal status were significant predictors for malignancy underestimation. The final pathological results of the underestimated NBs, distributed by lesion type were as follows: For FEA, we found 7 low grade DCIS, 1 high grade DCIS and 1 invasive tubular carcinoma. Among 20 cases of ADH, we encountered 15 low grade DCIS, 1 invasive tubular carcinoma, and 4 NST ductal carcinoma. In 5 cases originally diagnosed as ALH we upgraded to LCIS, we found 1 invasive lobular carcinoma, and 1 invasive ductal carcinoma. Finally, both of the combined lesions resulted in upgrade to invasive ductal carcinomas.
CONCLUSION: In our experience, we found that the diagnostic underestimation rate when using CNB is approximately three times than for VABB. Residual microcalcifications and presence of ALH on NB were found to be independent predictors of underestimation. In our country, surgical excision of FEA, ADH and ALH remains the standard of care. Further research is needed in order to establish which patients with atypical findings on initial biopsy will achieve benefit from surgery, and which might be suitable candidates for surveillance.
Citation Format: Calvo MF, Allemand C, Valerio AC, Hernandez MN, Corrao FH, Castro Barba M, Wernicke A, Orti R, Ilzarbe F, Gogorza SJ, Albrecht A, Izbizky GH, Lorusso C. Prediction of underestimation associated with flat epithelial atypia, atypical ductal hyperplasia and atypical lobular hyperplasia by needle biopsy: Experience in an Argentine breast unit [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P3-13-09.
