For mammography, it is necessary to complete diagnostic imaging evaluation before a probably benign assessment is assigned. A BI-RADS 3 assessment for mammograms is appropriate for a solitary circumscribed mass that appears solid at US, a focal asymmetry with no sonographic correlate, or a group of calcifications appearing punctate on magnification views. It is appropriate for these findings at baseline mammogram or without prior comparison but NOT for new findings on subsequent mammograms. A recent study evaluating the outcomes of BI-RADS 3 lesions on mammograms noted that most cancers (73%) were detected at the 6- or 12-month diagnostic study, and only 1/15 (6.7%) cancers were detected at 18-month follow-up. This validated necessity of short-interval follow-up of mammographic BI-RADS category 3 findings since 57.8% malignancies were diagnosed at or before 6 months. Furthermore, results from the National Mammography Database showed that BI-RADS 3 cancer yield overall was <2% (1.86% cumulative cancer yield through 2-year follow-up). When staging was available, 91.4% of cancers were stage 0 or I with ductal carcinoma in situ overrepresented, without evidence of advanced stage due to the 6–24-month delay in diagnosis. Data however did also find that cancer yield exceeds the accepted 2% threshold for patients older than 60 years (specifically: 2.8% for women in their 60s, 3.1% for women in their 70s, and 4.6% for women aged 80–90 years). However, there was no greater rate of advanced cancer staging of BR3 findings in older women. Therefore, recommendation is that surveillance may still be reasonable in an older population after having a informed discussion with the patient of the risks and benefits.
Historically, ultrasound Bi-RADS 3 criteria include: Solid mass (circumscribed margin, oval shape, and parallel orientation), Isolated complicated cyst, or Clustered microcysts. Other findings may be based on expert opinion including fat necrosis, architectural distortion thought to be due to postsurgical scar, or refraction shadowing at the edges of fat lobules. However, many findings historically assessed as BI-RADS 3, probably benign, now have sufficient data validating a BI-RADS 2, benign assessment. These include complicated cysts with debris, clustered microcysts, and multiple bilateral circumscribed masses.
BI-RADS 3 assessments can be made on screening ultrasound. Many practices obtain a full set of diagnostic images using hand-held ultrasound by technologist and then recalling any suspicious findings. Data shows that 0.6% - 20% of women receive BI-RADS 3 at the time of initial screening US. Correlation with mammography is mandatory when performing screening US as it can avoid US recall if the finding is mammographically stable and it can also further characterize the mammographic finding.
A BI-RADS 3 assessment with 12-month follow-up is reasonable for apparently solid, circumscribed masses on baseline screening hand-held US and follow-up can be performed as a screening examination provided the mass is redocumented. While several studies report that new BI-RADS 3 masses detected on incidence screening US can also be followed at 12 months, greater caution is needed in that setting, and particularly if there is a corresponding new mass on mammography in a postmenopausal woman. Biopsy is appropriate for lesions that demonstrate a growth in diameter >20% in 6 months (annualized rate of 44%), in women with synchronous breast cancer, or under certain clinical circumstances such as pregnancy.
For MRI, there are two findings that can be considered for a BI-RADS 3 assessment: a solitary enhancing focus separate from background without a high T2 correlate on a baseline examination, or a circumscribed mass on a baseline examination with benign characteristics. In terms of frequency of BI-RADS 3 assessments, the goal is less than 10%, with ideally a rate of approximately 1–2%. However, studies show that there is great practice variability in the use of BI-RADS 3 in MRI, with frequencies of use range from 6% to 24% and malignancy rates from 0% to 10%.
It is recommended to avoid BI-RADS 3 follow-up for benign parenchymal enhancement (BPE), which ideally should be classified as benign BPE (BI-RADS 1), or suspicious non–mass enhancement (NME) (BI-RADS 4 or 5). If the BPE is felt to be due to hormonal change, recommendation is for a BI-RADS 3 assessment with a short-term follow-up, with repeat examination during the optimal time of the menstrual cycle (week 2) or several weeks after stopping hormone replacement therapy. The pooled malignancy-rate of BI-RADS 3 lesions in MRI is 1.6% (95%-CI:0.9-2.3%), with highest malignancy rates in nonmass enhancement (NME), followed by mass, and then by foci. On high risk screening MRI, breast imagers should exercise greater caution in using BI-RADS 3 on nonbaseline examinations.
MR-directed ultrasound can help upgrade or downgrade BI-RADS 3 findings on MRI. Second-look (MRI directed) ultrasound has potential of detection of up to 20.7% malignant lesions. Patients may benefit from additional support and resources to help increase overall compliance and timely compliance of MRI follow-up.