Is stereotactic large-core needle biopsy beneficial prior to surgical treatment in BI-RADS 5 lesions?

摘要

INTRODUCTION: Due to screening mammography, more nonpalpable mammographic lesions warrant histological evaluation. Stereotactic large-core needle biopsy (SLCNB) has been shown to be as effective in diagnosing these lesions as diagnostic surgical excision, and has become the preferred diagnostic procedure for most mammographic lesions. Since radiologically malignant BI-RADS 5 lesions are almost always carcinoma, some centers advocate prompt diagnostic surgical excision for these lesions instead of SLCNB. For some patients this diagnostic surgical intervention may serve as definitive treatment. We set out to find a subgroup of mammographic BI-RADS 5 lesions for which surgical biopsy might be preferable. METHODS: Of 1644 consecutive nonpalpable lesions referred for SLCNB between April 1997 and May 2002, 238 were classified as BI-RADS 5. We assessed the number of carcinomas and the surgical interventions performed. Outcomes were compared between various types of mammographic lesions: density with calcifications, density without calcifications, and calcifications only. Different theoretical strategies for diagnostic work-up of BI-RADS 5 lesions were explored. RESULTS: Carcinoma was found in 229/238 lesions (96%). Most mammographic densities were invasive cancer (97%), while calcifications only showed the highest risk for DCIS (51%). In our study (current practice) all lesions were scheduled to first undergo SLCNB. A scenario was proposed where all lesions with only a density would be scheduled directly for sentinel node biopsy (SNB) and tumour excision (n = 154; 65%), while other lesions would still be scheduled for SLCNB. When we compared this scenario to current practice, four out of 238 patients (< 2%) would be 'overtreated' with SNB. CONCLUSIONS: Our findings confirm a high predictive value of malignancy for BI-RADS 5 lesions (96%). Surgical excision is therefore imperative for all BI-RADS 5 lesions, irrespective of SLCNB results. For BI-RADS 5 lesions presenting as mammographic densities only, we propose to consider surgical excision with SNB to be the first diagnostic and therapeutic procedure. SLCNB is preferred in all other cases.