Date of Award


Document Type


Degree Name

Medical Doctor (MD)

First Advisor

Liane Philpotts


Although core needle biopsy has been shown to be effective in diagnosing both benign and malignant mammographically detected lesions in the breast, it has also been shown to underestimate cancer most likely due to sampling error. Since a diagnosis of atypical hyperplasia versus malignancy is based on quantitative factors (which could be affected by an error in sampling), the current recommendation is surgical excision for atypical hyperplasia diagnosed on core biopsy. The purpose of the study was to determine if a subset of patients with atypia diagnosed by core biopsy fit the Breast Imaging Reporting and Data Systems (BI-RADS) Category 3, probably benign, definition of having a less than 2% chance of being carcinoma at subsequent surgical excision when comparing histologic subtype, mammographic findings, core biopsy factors, and clinical factors. For this subset of patients, imaging follow-up, rather than surgical excision could be recommended. Retrospective searches of the breast imaging and pathology databases from 1992 to August 2005 were performed to identify all cases of atypia found on core biopsy. The data collection and database use were HIPAA-compliant and followed the protocols of the institutional review board. The pathology reports were reviewed to determine the histologic type: atypical ductal hyperplasia (ADH), atypical lobular hyperplasia (ALH), mixed, or other atypia. The ADHs were further classified as to focal/mild, not otherwise stated (NOS), or marked based on the pathology reports. Follow-up information was obtained to identify cases in which lesions that were initially diagnosed as atypia at the time of core biopsy were later upgraded to malignancy after subsequent surgical excision or mammographic follow-up. The histologic subtype, mammographic findings, core biopsy factors, and clinical factors were compared to lesions which were not upgraded to carcinoma. The results were analyzed with a Chi-square test, with p< 0.05 indicative of significant difference. There were 327 cases of atypia found in the 3898 (8%) core needle biopsies that were performed during the above stated time period. The histologic subtypes were: ADH (75%), ALH (13%), mixed (4%), other (7%). There was an overall malignancy rate of 13%. Malignancy was found in 14% of ADH lesions, 5% of ALH, 20% of mixed, and 10% of other atypias on excision. The 215 ADH cases were further examined in their histologic subtypes (37% were focal/mild, 42% NOS, and 20% ADH marked). Malignancy was found in 6% focal/mild ADH, 10% NOS ADH, and 40% ADH marked. When comparing all the factors considered, the lowest underestimation rate (3%) was found in patients with focal/mild ADH diagnosed with a vacuum- assisted 11-gauge biopsy needle. Upgrade rates vary significantly depending on classification of ADH and the type of atypia. While severe forms of atypia (NOS ADH, ADH marked, mixed atypias, and other atypias) should continue to receive routine surgical excision, there are selected subsets of patients with whom other management options could possibly be considered. For patients with focal/ mild atypical ductal hyperplasia diagnosed at the time of core biopsy with a vacuum-assisted 11-gauge needle, imaging follow-up (mammography or MRI) could be considered on an individual basis.