Diagnostic Performance Of Ultrasound-guided Core Needle Biopsy In Breast Lesions: A Clinical Study

Objective:1. To assess the accuracy of ultrasound-guided16G and18G core needle biopsy（CNB）for ultrasound visible breast lesions with different sonographic features.2. To evaluate underestimate rates of breast lesions diagnosed at ultrasound-guided breastbiopsy.Methods:1Between July2005and July2012,4,453ultrasound-detected breast lesions underwentultrasound-guided CNB and were retrospectively reviewed. Surgical excision wasperformed for955lesions. Histological findings were compared between theultrasound-guided CNB and the surgical excision to determine sensitivity,false-negative rate, agreement rate, according to different lesion features.2(1)4,453ultrasound-guided visible breast lesions were respectively reviewed from2005.7to2012.7. In the955lesions which were under surgical excisions, histologicalfindings were compared between the ultrasound-guided CNB and the surgical excisionto analysis the underestimate rates and reasons for underestimation.(2) In all the ultrasound-guidedbreast biopsies, totally207papillary lesions werefound.90of these lesions underwent surgical excision,110had a minimum of1yearimaging follow-up, and7were lost. Histological findings of the ultrasound-guidedbiopsy and surgical excision were compared for underestimate rates according toBI-RADS category and biopsy methods.Results:1. In total，there are566biopsieswere performed with16G CNB and389with18G CNB.Final pathological results were malignant in84.1%, high-risk in8.4%, and benign in7.5%. False-negative rates were1.4%for16G and18G CNB. The overall agreement rates were92.4%for16G CNB and92.8%for18G CNB; the complete sensitivities、absolute sensitivities and false negative rates were both98.6%,93.9%and1.4%，respectively; Meanwhile, for both16G and18G CNB, the agreements were better formass lesions (16G:92.7%;18G:93.7%) than non-mass lesions (16G,85.7%;18G,78.3%)(P<0.01in18G-CNB). For the mass lesion groups with the diameter equal toor less than10mm, the agreement rates (16G,83.3%;18G,86.7%) were lower thanthe overall data(P<0.01). There was no statistical significance between groups withand without calcification (P>0.01).2.(1)123underestimate lesions were found in total. The high-risk underestimate ratesand DCIS underestimate rates were48.0%and46.2%for16G CNB VS53.3%and41.2%for18G CNB; there was no statistical significance between the twogroups(P>0.01).52.5%of the high risk underestimationwasADH underestimation.(2) A total of29lesions were underestimated; the underestimate rate of benignintraductal papilloma(IDP) with concordance imaging-histologic findings was5.8%;whereas the underestimate rate of IDP with diacordant imaging-histologic finding was40%;47.8%papilloma with atypical ductal hyperplasia(IDP+ADH) were upgraded tomalignant. In total,16.0%understimation were biopsied by16G core needlebiopsy(CNB) and17.9%by18G CNB, which were significantly higher thanvacuum-assisted biopsy(VAB)(P<0.05).Conclusion:1. Ultrasound-guided16G and18G CNB are accurate for evaluating ultrasound-visiblebreast mass lesions with a diameter>10mm.2.(1) The underestimate rates of high risk lesion and DCIS diagnosed by US-guided16Gor18G CNB are relative high. VABor surgical excisions should be performed.(2)For high underestimations of IDP+ADH and IDP with discordant imaging-histologicfindings, more accuratemethod than16G or18G CNB，as VAB, or surgical excisionsshould be performed. Concordant IDP can be followed up.