Histopathological Upgrade in Papillary Breast Lesions Diagnosed on Core Needle Biopsy: Predictors and Clinical Implications from a Tertiary Care Cohort: Upgrade in Papillary Breast Lesions

Background: Papillary breast lesions encompass a heterogeneous spectrum, and core needle biopsy (CNB) may underestimate in situ or invasive malignancy. We evaluated malignant upgrade and preoperative factors associated with upgrade after surgical excision.Methods: This retrospective cohort study identified 298 patient records with papillary breast lesions. A total of 14 patients were excluded: eight who lacked a preceding CNB, five who did not undergo subsequent surgical excision, and one whose index papillary diagnosis and surgery both occurred before the January 2019–December 2025 study period. This left 284 patients for analysis. Malignant upgrade was defined as ductal carcinoma in situ or invasive malignancy on excision after a benign or atypical papillary lesion on CNB; atypical ductal hyperplasia alone was not considered malignant upgrade. Firth-penalized logistic regression was used to obtain adjusted estimates because a highly predictive atypia variable was present.Results: Malignant upgrade occurred in 137 of 284 patients (48.2%). Upgrade was observed in 130 of 150 lesions with atypia (86.7%) and 7 of 134 lesions without atypia (5.2%). In adjusted analysis (n = 273 complete cases), atypia on CNB was the strongest independent predictor (adjusted odds ratio [aOR] 71.49, 95% CI 25.96–196.84; P < .001). Increasing age (aOR 1.04 per year, 95% CI 1.01–1.07; P = .02) and manual rather than ultrasound-guided core biopsy (aOR 5.01, 95% CI 1.46–17.20; P = .010) also remained associated with malignant upgrade. Ultrasound lesion size, palpability, peripheral location, and BI-RADS ≥4 were not independently associated with malignant upgrade after adjustment.Conclusion: Atypia on CNB was associated with a very high probability of malignant upgrade and supports surgical excision. The low upgrade rate among lesions without atypia is hypothesis-generating but cannot establish the safety of surveillance because only surgically excised lesions were analyzed.