Introduction

This retrospective study aimed to determine the percentage of MRI-detected breast cancers diagnosed using targeted ultrasound and standard 14-gauge (14g) biopsy in the setting of an Australian breast MRI service. This is of clinical relevance because malignancies not identifiable on mammography or by ultrasound may then require more invasive, technically demanding and costly MRI-guided interventional procedures, usually by large-core vacuum-assisted biopsy (VAB) or hook-wire localisation with open surgical biopsy.

Methods

On review of the 12-year period 2006–2018, we identified 67 new breast cancer events in 64 women where the diagnosis was made on the basis of the MRI scan findings either alone (n = 60) or in combination with a concurrent mammogram (n = 7), with no recorded clinical abnormality. The percentage in which malignancy was confirmed on histopathology using targeted ultrasound in combination with 14g biopsy was determined versus other biopsy methods, for both invasive cancer and in situ disease.

Results

Ultrasound-guided 14g biopsy was successful in establishing the presence of malignancy in 42/46 (91%) of events with a final diagnosis of invasive cancer, with 2 more proven by MRI-guided interventional procedures (1 VAB and 1 hook-wire) and 1 by open surgical biopsy. In the final case, a 5 mm focus on MRI with no sonographic correlate at the initial presentation was only identified and biopsied by ultrasound at 12-month follow-up.

For events with a final diagnosis of DCIS/pLCIS (pleomorphic LCIS, n = 2), US-guided 14g biopsy was successful in 10/21 (48%), while 4/7 events with corresponding mammographic microcalcifications were proven by x-ray stereotactic interventions. A further 5 events had MRI-guided interventions (3 VAB and 2 hook-wires) and 1 an open surgical biopsy to confirm malignancy. In the final case (a woman with a 30 × 20 mm focal area of non-mass enhancement with corresponding microcalcifications consistent with DCIS), a pathologic diagnosis was not made until the patient presented 5 years later with invasive disease. There were also 3 instances of upgrades to invasion on final surgical pathology, one from pLCIS to microinvasion and 2 others from DCIS to IDC. Among the DCIS/pLCIS events, semi-random 14g core biopsy (sampling at the expected location of the MRI abnormality without a specific sonographic correlate) proved to be successful in 3 women.

Conclusion

Ultrasound-guided 14g core biopsy established a malignant diagnosis in 91% of invasive cancers and in 48% of DCIS/pLCIS cases. This relatively non-invasive, technically easy to perform and low-cost biopsy procedure can be used immediately when targeted ultrasound shows a correlate for a suspicious MRI scan finding. Careful imaging-pathologic correlation is required after 14g biopsy, and a discordant result will usually prompt recourse to an MRI-guided VAB or hook-wire localisation.