Surgical upgrade rate of breast atypia to malignancy: An academic center's experience and validation of a predictive model.

Atypical ductal hyperplasia (ADH), atypical lobular hyperplasia (ALH), and lobular carcinoma in situ (LCIS) are commonly seen on breast core needle biopsy (CNB). Many institutions recommend excision of these lesions to exclude malignancy. A retrospective chart review was performed on patients who had ADH, ALH, or LCIS on breast CNB from 1/1/08 to 12/31/10 who subsequently had surgical excision of the biopsy site. Study objectives included determining upgrade to malignancy at surgical excision, identification of predictors of upgrade, and validation of a recently published predictive model. Clinical and demographic factors, pathology, characteristics of the biopsy procedure and visible residual lesion were recorded. T test and chi-squared test were used to identify predictors. Classification tree was used to predict upgrade. 151 patients had mean age of 53 years. The mean maximum lesion size on imaging was 11 mm. The primary atypia was ADH in 63.6%, ALH in 27.8%, and LCIS in 8.6%. 16.6% of patients had upgrade to malignancy, with 72% DCIS and 28% invasive carcinoma. Risk factors for upgrade included maximum lesion size (P = .002) and radiographic presence of residual lesion (P = .001). A predictive model based on these factors had sensitivity 78%, specificity 80% and AUC = 0.88. Validating a published nomogram with our data produced accuracy figures (AUC = 0.65) within published CI of 0.63-0.82. In CNB specimens containing ADH, ALH, or LCIS, initial lesion size and presence of residual lesion are predictors of upgrade to malignancy. A validated model may be helpful in developing patient management strategies.