Abstract: Objective To assess the accuracy of ultrasound-guided 16G and 18G core needle biopsy for detecting ultrasound
visible breast lesions with different sonographic features. Methods A total of 955 sonographically detected breast lesions
examined with ultrasound-guided core needle biopsy (US-CNB) and subsequently surgically excised from July 2005 to July
2012 were retrospectively reviewed. Histological findings of US-CNB and the surgical specimens were analyzed for
agreements, sensitivities, false negative rates, and underestimate rates according to different sonographic features. Results The
pathological results of the US-CNB showed malignant lesions in 84.1%, high-risk lesions in 8.4%, and benign lesions in 7.5% of
the samples. The overall agreement rates were 92.4% for 16G CNB and 92.8% for 18G CNB; their complete sensitivities and
false negative rates were both 98.6% and 1.4%, respectively; the high-risk underestimate rates and DCIS underestimate rates
were 48.0% and 46.2% for 16G CNB vs 53.3% and 41.2% for 18G CNB, showing no significant difference between the two
groups (P>0.01). For both 16G and 18G CNB, the agreements were better for mass lesions than for non-mass lesions (P<0.01).
For the mass lesions with a diameter no greater than 10 mm, the agreement rates were lower than the overall data (P<0.01).
Calcification in the lesions did not affect the agreement rates (P>0.01). Conclusion Ultrasound-guided 16G and 18G CNB are
both accurate methods for evaluating ultrasound visible breast mass lesions with a diameter larger than 10 mm.