Multiple Bilateral Circumscribed Masses at Screening Breast US: Consider Annual Follow-up

Multiple bilateral circumscribed benign-appearing masses are common at screening US, and there were no malignancies among 127 lesions with at least 2 years of follow-up (95% confidence interval: 0%, 2.9%) in this prospective series.

Abstract

Purpose:

To determine prospectively the prevalence and rate of malignancy of multiple bilateral (MB) circumscribed breast masses detected at screening ultrasonography (US) compared with those of other US-depicted masses.

Materials and Methods:

This institutional review board–approved, HIPAA-compliant prospective trial included women at elevated risk for breast cancer, who gave written informed consent to participate in a study evaluating cancer detection rates for three rounds of annual supplemental screening US at 21 international sites. After exclusions, 2662 participants and 7473 screening studies were included. Physician-performed US studies were interpreted, with blinding to mammography results. Simple cysts were noted. Breast Imaging Reporting and Data System features of all other findings were recorded, with addition of the descriptor MB similar-appearing circumscribed masses (minimum of three total and at least one in each breast), with details of the largest such mass recorded. Rates of malignancy were determined after biopsy or mammographic and US follow-up at a minimum of 11 months. For this analysis, 490 women (1370 screenings) with prior mastectomy were excluded. Descriptive statistics and exact 95% confidence intervals (CIs) were generated.

Results:

Of 2172 evaluable participants (6103 screening studies; median age at study entry, 54.0 years; range, 25–91 years), 1454 had unique findings at US. One hundred thirty-five (6.2%) participants had 153 unique MB circumscribed masses, with no malignancies (0% [95% CI: 0%, 2.4%]; 95% CI: 0%, 2.9% for the 127 masses with at least 2 years of follow-up). There were 1319 (60.7%) participants with 2464 non-MB lesions, including 1038 solitary circumscribed masses with a malignancy rate of 0.8% (eight of 1038). Of 836 solitary circumscribed masses with at least 2 years of follow-up, the malignancy rate was 0.4% (three of 836; 95% CI: 0.1%, 1.0%). Of the 135 women with MB circumscribed masses, 82 (60.7%) also had a solitary lesion. Two of these 82 women (2.4%) had cancer.

Conclusion:

MB similar-appearing circumscribed masses seen at screening US are almost always benign, with no malignancies found among such lesions in this prospective, multicenter experience. These lesions are suitable for diagnostic follow-up in 1 year, with resumption of screening thereafter if they are stable.

© RSNA, 2013

Introduction

Multiple bilateral circumscribed masses have been reported in 1.7% of screening mammography studies and can be classified as Breast Imaging Reporting and Data System (BI-RADS) category 2, benign findings, without the need for recall for further evaluation (1). Many such masses represent cysts and/or complicated cysts, which often show a fluctuating pattern, with some cysts developing and others regressing. From 10% to 15% of fibroadenomas are multiple, and, when fibroadenomas are multiple, they are often bilateral (2). Among 1440 women with multiple bilateral mammographic masses reported by Leung and Sickles (1), two (0.14%) node-negative interval cancers were found—less than the expected interval cancer rate.

The American College of Radiology Imaging Network (ACRIN) protocol 6666 evaluated annual physician-performed screening whole-breast ultrasonography (US) in women with dense breasts and at least one other risk factor (3). In the ACRIN 6666 protocol, all US findings were prospectively described by using the BI-RADS US categories (4), with the addition of the descriptor “multiple bilateral circumscribed masses.” At US, such masses include complicated cysts, clustered microcysts, and solid masses. Simple cysts were considered separately, as they are a typically benign finding at US. The purpose of our analysis was to determine prospectively the prevalence and rate of malignancy of multiple bilateral circumscribed masses detected at screening US compared with those of other sonographically depicted masses.

Materials and Methods

Institutional Review Board Approval, Consent, and Data Monitoring

The study was compliant with the Health Insurance Portability and Accountability Act and received institutional review board approval from all participating sites, as well as approval from the ACRIN and the National Cancer Institute Cancer Imaging Program. The ACRIN Data and Safety Monitoring Committee conducted a review every 6 months while the study was open to accrual. Participants provided written informed consent. Web-based data capture and quality monitoring were performed by the ACRIN Biostatistics and Data Management Center. Site data quality was audited periodically for protocol compliance and completeness by personnel from ACRIN headquarters; W.A.B. and E.B.M. performed US image quality review of randomly selected cases and providedfeedback to sites during the first 18 months of the study.

Investigators

All radiologist investigators were Mammography Quality Standards Act–qualified interpreting radiologists who met trial-specific experience requirements, including experience in interpreting 2500 mammographic examinations and performing and interpreting at least 500 breast US examinations, each within the previous 2 years (5). All investigators successfully completed qualification tasks in both US scanning (6) and mammographic and sonographic interpretive skills (7).

Participants

Participants were asymptomatic women with at least heterogeneously dense breast tissue (8) in at least one quadrant and at least one other risk factor for breast cancer (3,9) who presented for routine annual mammography (ie, they were asymptomatic, and routine follow-up had been recommended on the basis of their most recent previous mammographic study). Race and ethnic group were self-assigned from a list of options for ethnicity and a series of “yes” or “no” questions for race, and women could belong to more than one category. A total of 2809 women were recruited from 21 sites in the United States, Canada, and Argentina between April 2004 and February 2006. After ineligible participants (n = 84), those with missing examinations (n = 11), and those with missing biopsy or follow-up results or other exclusion criteria for a given examination were excluded (as detailed in Berg et al [3]), data for 2662 participants and 7473 screening examinations remained evaluable. Women aged at least 25 years who presented for routine annual mammography were eligible to participate if they met study definitions of elevated risk (3) as determined by research personnel and if they had heterogeneously dense or extremely dense parenchyma (8) in at least one quadrant, according to either a prior mammography report or a review of prior mammographic studies. Women were excluded from enrollment if they were pregnant or lactating; if they were known to have metastatic disease; if they had signs or symptoms of breast cancer; if they had undergone breast needle biopsy or surgery within the previous 12 months; if they had undergone breast magnetic resonance (MR) imaging, bilateral whole-breast US, tomosynthesis, and/or contrast material–enhanced mammography or US within the previous 12 months; and/or if they had breast implants. We have previously reported the cancer detection rate with supplemental screening US each year for 3 years in the main study population (3). We have also previously reported on cystic breast lesions, including simple cysts, complicated cysts, and complex cystic and solid breast masses, without regard to multiplicity or bilaterality, in the original analysis set (10). Because only women with two breasts could have multiple bilateral findings, 490 participants with a personal history of breast cancer and unilateral mastectomy prior to study entry were excluded from this analysis, leaving 2172 women with 6103 screening examinations in this analysis set. One multiple bilateral benign lesion was described as irregular masses and was excluded from the multiple bilateral circumscribed lesion analysis set (and was grouped with suspicious findings). Six multiple bilateral lesions were calcifications without a mass and were grouped with other special cases, as defined by the BI-RADS US lexicon (4).

Screening Methods

Each participant underwent mammographic and physician-performed US screening examinations in randomized order, with the interpreting radiologist for each examination masked to the results of the other study, at times 0, 12, and 24 months, with clinical outcome recorded at 36 months (n = 1893); results of imaging at 36 months were also recorded when available (n = 1684). If either screening mammography or US resulted in a recommendation for anything other than routine screening, the test was considered positive, and an integrated interpretation was then performed by one of 100 qualified site investigators (including W.A.B. and E.B.M.), who reviewed the study mammographic and US examinations together and recorded results. Clinical management was based on recommendations from the integrated interpretation.

Mammography with at least two views was performed by using screen-film (n = 3558) or digital (n = 2545) technique. The visually estimated overall breast density on study mammograms was recorded as less than 25% dense, as 26%–40% dense, as 41%–60% dense; as 61%–80% dense, or as more than 80% dense. Qualified physician investigators performed US scanning within 2 weeks of the mammographic examination (and usually the same day) with standardized technique and documentation (9) and by using high-resolution linear-array broad-bandwidth transducers with a maximum frequency of at least 12 MHz. The outer breast was scanned with the participant in the supine oblique position with the ipsilateral arm raised, and the inner breast was scanned with the participant in the supine position. The largest simple cyst in each quadrant was documented. For lesions other than simple cysts, orthogonal images were obtained with and without calipers, with and without spatial compounding, and with and without power Doppler. For a negative examination, protocol required a minimum of one image per quadrant and one behind the nipple. The axilla was electively included by investigators in 2073 (34%) of 6103 examinations (723 women).

BI-RADS features (4,8) and assessments of each lesion and by breast were recorded. The expanded seven-point BI-RADS assessment scale was used for all imaging interpretations. On this scale, a score of 1 indicates a negative study; a score of 2, a benign finding; a score of 3, a probably benign finding; a score of 4a, low suspicion; a score of 4b, intermediate suspicion; a score of 4c, moderate suspicion; and a score of 5, a finding highly suggestive of malignancy. Interpretive criteria were prospectively defined (5). Simple cysts were recorded and were to be considered BI-RADS category 2, benign findings. According to protocol, all multiple bilateral circumscribed benign-appearing masses identified mammographically were to be assessed as benign; such masses seen only at US, including solid masses and complicated cysts with debris, were to be classified as probably benign (5). Each lesion had a consistent lesion number that was used across time points and on all biopsy and surgery forms. BI-RADS features, including the location, size, shape, and margins of the largest multiple bilateral mass in a given participant, were prospectively recorded, as was the characteristic that the finding was “multiple, bilateral, circumscribed masses.” For presumed-solid masses, we distinguished an oval shape from “two or three gentle lobulations,” even though the BI-RADS US lexicon (4) does not. Interpretive criteria at both mammography and US were standardized (5). Recommendations were recorded separately from BI-RADS assessments and included routine annual imaging, short-interval follow-up in 6 months, additional imaging, and biopsy. At the time of US imaging, investigators were asked to determine and record whether each finding was palpable.

Reference Standard

The reference standard was the most severe of biopsy results within 365 days of the mammographic screening or imaging and/or clinical follow-up at 1 year. One-year follow-up was targeted for each screening examination for 365 days after the previous mammographic screening examination. Of the 2172 women in this analysis, a subset of 526 underwent a screening MR imaging examination after the third annual screening with mammography and US, as previously described (11). We also considered the confidence interval (CI) for malignancy rates after excluding lesions lacking either biopsy or mammographic and US follow-up of at least 2 years.

A complete examination of all study breasts performed more than 11 full months after the previous screening examination was considered the next annual screening examination. Very few visits were early: In 2169 women in year 1, 22 of the second screening examinations occurred before 11 months, of which 11 occurred before 10.5 months; in 2041 women in year 2, 28 of the third screening examinations occurred before 11 months, of which 14 occurred before 10.5 months; and in 1893 women in year 3, 23 of the fourth screening examinations occurred before 11 months, of which 14 occurred before 10.5 months. The absence of a known diagnosis of cancer at a participant interview, review of medical records, or both at least 11 full months (330 days) after the mammographic screening examination was considered to indicate negativity for disease, as were six prophylactic mastectomies that revealed no evidence of cancer at pathologic examination. Biopsy results that showed breast cancer (in situ or infiltrating ductal carcinoma or infiltrating lobular carcinoma) in the breast or axillary lymph nodes were considered to indicate positivity for malignant disease. Three women with incomplete results at the first screening did complete the subsequent screenings and were included in the main analysis, for a total of 2172 participants (3). Of these 2172 women, 938 (43.2%) had a personal history of breast cancer. We analyzed malignancy rates for the 135 women with multiple bilateral circumscribed masses and the 672 women with solitary circumscribed benign-appearing masses. In addition, we analyzed malignancy rates for the 118 women with multiple bilateral circumscribed masses and the 561 women with solitary circumscribed masses who underwent biopsy or at least 2 years of follow-up.

Statistical Analysis

Descriptive summaries of the frequency of multiple bilateral masses, as well as other masses, were prepared, and rates of malignancy were determined. Malignancy rates were compared across mass types according to US BI-RADS features. The demographic data in participants with multiple bilateral masses at US were compared with data in participants without such masses by using the Pearson χ² test or two-sample Student t test. For this substudy, lesion-level details of malignant status were used. Lesions seen at multiple time points were considered to have the features at first description. Exact 95% CIs for the malignancy rates were calculated. The malignancy rates of the multiple bilateral circumscribed masses and those of the solitary benign-appearing masses were compared by using the Fisher exact test. P < .05 was considered to indicate a significant difference.

Results

Participant Demographic Data

A total of 6103 screening studies in 2172 participants were available for analysis. For 2079 (96%) participants, prior mammograms were available, and 867 (40%) had at least targeted prior US images available at the time of initial reporting; in years 2 and 3, prior study mammograms and whole-breast US images were available. Of 2172 participants, 1454 (66.9%) had a US finding other than a simple cyst at at least one of the three annual screening examinations. One hundred sixty-six unique multiple bilateral lesions were detailed, of which 153 qualified as multiple bilateral circumscribed masses in 135 women (representing 6.2% of the 2172 participants with two breasts [95% CI: 5.2%, 7.3%]) (Table 1). Eighty-two of the 135 women with multiple bilateral findings also had another US finding (Figure). Women with multiple bilateral circumscribed lesions were younger, with a median age of 50 years (mean, 51.3 years; range, 32–77 years), than women with other findings at US (median age, 55 years; mean, 55.3 years; range, 25–85 years; P < .0001). Multiple bilateral findings were more common in pre- and perimenopausal women (72 [9.8%] of 737) than in postmenopausal women (45 [4.4%] of 1030; P < .0001). Multiple bilateral findings were less common in women with a personal history of breast cancer (21 [2.2%] of 938) than in women without a personal history of breast cancer (114 [9.2%] of 1234; P < .0001). Only 91 (9.7%) of the 938 women with a personal history of breast cancer (and two breasts) had cysts in the remaining previously affected breast, compared with 621 (50.3%) of the 1234 other participants (P < .001). Approximately 6.3% (128 of 2033) of white women and 6.1% (four of 66) of Asian women had multiple bilateral lesions, compared with 3.6% (eight of 225) of Hispanic women and 3.2% (two of 63) of African American women. Increasing breast density correlated with a higher incidence of multiple bilateral lesions at US—ranging from 2.0% for women with breast density of less than 40% (at visual evaluation) to 10.5% for those with breast density greater than 80% (Table 1).

Table 1.

Demographic Data in Participants with and Those without Multiple Bilateral Circumscribed Benign-appearing Masses at Screening US over Three Annual Examinations

Note.—Unless otherwise specified, data are numbers of participants, with percentages in parentheses. For these data, only participants with two breasts were considered. Multiple bilateral masses = at least three circumscribed benign-appearing masses (excluding simple cysts) were present, two in one breast and one in the other. ADH = atypical ductal hyperplasia, ALH = atypical lobular hyperplasia, LCIS = lobular carcinoma in situ.

^*Data are medians, with ranges in parentheses.

^†Regardless of other risk factors.

^‡Breast density was visually assessed at the time of each annual mammographic examination.

^§Participants with multiple risk factors were considered to have a primary risk factor by using the following hierarchy: mutation in BRCA1 or BRCA2; history of chest, mediastinal, and/or axillary irradiation; personal history of breast cancer; Gail (12) or Claus (13) model lifetime risk ≥ 25%; Gail model 5-year risk ≥ 2.5%; Gail model 5-year risk ≥ 1.7% and extremely dense breasts; and prior biopsy that revealed atypical ductal hyperplasia, atypical lobular hyperplasia, lobular carcinoma in situ, or atypical papilloma.

Figure a:

Images in 56-year-old woman with multiple similar bilateral masses at US and cancer. This woman had a lifetime risk for breast cancer of more than 25%. (a) Craniocaudal (left) and mediolateral oblique (right) mammograms demonstrate heterogeneously dense parenchyma. (b) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 2 cm from the nipple show an oval circumscribed isoechoic benign-appearing mass (arrows). (c) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 6 cm from the nipple show a similar oval circumscribed benign-appearing hypoechoic mass with tiny eccentric cystic foci (arrows). (d) Radial (left) and antiradial (right) US images in the 1:30-o’clock position in the left breast 2 cm from the nipple show an oval circumscribed isoechoic mass (arrows) consistent with benign etiology. Multiple bilateral similar benign-appearing masses (with at least two in one breast and one in the other) can be considered a benign finding at screening US. (e) As shown on radial (left) and antiradial (right) US images, this woman also had a suspicious irregular isoechoic mass (arrows) in the left breast in the 11-o’clock position 4 cm from the nipple. US-guided biopsy and excision showed a 9-mm mass due to grade 2 invasive and intraductal carcinoma, with a 2-mm invasive component and negative findings at sentinel node biopsy. This case illustrates the importance of careful scrutiny of each finding.

Figure b:

Images in 56-year-old woman with multiple similar bilateral masses at US and cancer. This woman had a lifetime risk for breast cancer of more than 25%. (a) Craniocaudal (left) and mediolateral oblique (right) mammograms demonstrate heterogeneously dense parenchyma. (b) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 2 cm from the nipple show an oval circumscribed isoechoic benign-appearing mass (arrows). (c) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 6 cm from the nipple show a similar oval circumscribed benign-appearing hypoechoic mass with tiny eccentric cystic foci (arrows). (d) Radial (left) and antiradial (right) US images in the 1:30-o’clock position in the left breast 2 cm from the nipple show an oval circumscribed isoechoic mass (arrows) consistent with benign etiology. Multiple bilateral similar benign-appearing masses (with at least two in one breast and one in the other) can be considered a benign finding at screening US. (e) As shown on radial (left) and antiradial (right) US images, this woman also had a suspicious irregular isoechoic mass (arrows) in the left breast in the 11-o’clock position 4 cm from the nipple. US-guided biopsy and excision showed a 9-mm mass due to grade 2 invasive and intraductal carcinoma, with a 2-mm invasive component and negative findings at sentinel node biopsy. This case illustrates the importance of careful scrutiny of each finding.

Figure c:

Images in 56-year-old woman with multiple similar bilateral masses at US and cancer. This woman had a lifetime risk for breast cancer of more than 25%. (a) Craniocaudal (left) and mediolateral oblique (right) mammograms demonstrate heterogeneously dense parenchyma. (b) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 2 cm from the nipple show an oval circumscribed isoechoic benign-appearing mass (arrows). (c) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 6 cm from the nipple show a similar oval circumscribed benign-appearing hypoechoic mass with tiny eccentric cystic foci (arrows). (d) Radial (left) and antiradial (right) US images in the 1:30-o’clock position in the left breast 2 cm from the nipple show an oval circumscribed isoechoic mass (arrows) consistent with benign etiology. Multiple bilateral similar benign-appearing masses (with at least two in one breast and one in the other) can be considered a benign finding at screening US. (e) As shown on radial (left) and antiradial (right) US images, this woman also had a suspicious irregular isoechoic mass (arrows) in the left breast in the 11-o’clock position 4 cm from the nipple. US-guided biopsy and excision showed a 9-mm mass due to grade 2 invasive and intraductal carcinoma, with a 2-mm invasive component and negative findings at sentinel node biopsy. This case illustrates the importance of careful scrutiny of each finding.

Figure d:

Images in 56-year-old woman with multiple similar bilateral masses at US and cancer. This woman had a lifetime risk for breast cancer of more than 25%. (a) Craniocaudal (left) and mediolateral oblique (right) mammograms demonstrate heterogeneously dense parenchyma. (b) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 2 cm from the nipple show an oval circumscribed isoechoic benign-appearing mass (arrows). (c) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 6 cm from the nipple show a similar oval circumscribed benign-appearing hypoechoic mass with tiny eccentric cystic foci (arrows). (d) Radial (left) and antiradial (right) US images in the 1:30-o’clock position in the left breast 2 cm from the nipple show an oval circumscribed isoechoic mass (arrows) consistent with benign etiology. Multiple bilateral similar benign-appearing masses (with at least two in one breast and one in the other) can be considered a benign finding at screening US. (e) As shown on radial (left) and antiradial (right) US images, this woman also had a suspicious irregular isoechoic mass (arrows) in the left breast in the 11-o’clock position 4 cm from the nipple. US-guided biopsy and excision showed a 9-mm mass due to grade 2 invasive and intraductal carcinoma, with a 2-mm invasive component and negative findings at sentinel node biopsy. This case illustrates the importance of careful scrutiny of each finding.

Figure e:

Images in 56-year-old woman with multiple similar bilateral masses at US and cancer. This woman had a lifetime risk for breast cancer of more than 25%. (a) Craniocaudal (left) and mediolateral oblique (right) mammograms demonstrate heterogeneously dense parenchyma. (b) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 2 cm from the nipple show an oval circumscribed isoechoic benign-appearing mass (arrows). (c) Radial (left) and antiradial (right) US images in the 10-o’clock position in the right breast 6 cm from the nipple show a similar oval circumscribed benign-appearing hypoechoic mass with tiny eccentric cystic foci (arrows). (d) Radial (left) and antiradial (right) US images in the 1:30-o’clock position in the left breast 2 cm from the nipple show an oval circumscribed isoechoic mass (arrows) consistent with benign etiology. Multiple bilateral similar benign-appearing masses (with at least two in one breast and one in the other) can be considered a benign finding at screening US. (e) As shown on radial (left) and antiradial (right) US images, this woman also had a suspicious irregular isoechoic mass (arrows) in the left breast in the 11-o’clock position 4 cm from the nipple. US-guided biopsy and excision showed a 9-mm mass due to grade 2 invasive and intraductal carcinoma, with a 2-mm invasive component and negative findings at sentinel node biopsy. This case illustrates the importance of careful scrutiny of each finding.

US Findings

The majority of both multiple bilateral circumscribed masses and other US findings were seen at the first screening round (Table 2). Of the 153 multiple bilateral circumscribed benign-appearing masses, 127 (83.0%) had at least 24 months of follow-up and the rest had at least 11 months of follow-up. Among the 153 multiple bilateral circumscribed benign-appearing masses, 98 (64.1%) were described as complicated cysts with debris. Another 50 of the multiple bilateral findings at US (in 44 women, 2.0% of all participants) were described as oval (n = 43) or gently lobulated (n = 7) circumscribed solid masses and another five were described as clustered microcysts (Table 3).

Table 2.

Timing of First Detection of US Lesions in 2172 Participants Screened with US at 0, 12, and 24 Months

Note.—Data are numbers of lesions, with percentages in parentheses.

^*One year of follow-up was defined as at least 11 full months of clinical and/or imaging follow-up or biopsy.

Table 3.

Malignancy Rates among Solitary Circumscribed and Multiple Bilateral Circumscribed Benign-appearing Masses Seen at 6103 Screening US Examinations in 2172 Participants

Note.—Data in parentheses are raw data. Data in square brackets are 95% CIs. NA = not applicable (no entries).

^*One lesion that was initially described as clustered microcysts at the first examination was described as a complex hypoechoic mass with tiny cystic areas at 12 months and may or may not have been the same lesion, an 18-mm node-negative invasive lobular carcinoma.

^†The total with exclusion of lesions with less than 2-year follow-up and no biopsy was 0.4% (three of 836; 95% CI: 0.1%, 1.0%) for solitary lesions and 0% (zero of 127; 95% CI: 0%, 2.9%) for multiple bilateral lesions.

Among 1319 women with other US findings, 2464 lesions were detailed, including 288 complicated cysts (in 221 women, 10.2% of the 2172 participants). Another 577 masses were described as oval (n = 437) or gently lobulated (n = 140) circumscribed masses (in 415 women, 19.1% of participants) (Table 3).

Malignancy Rates for Multiple Bilateral versus Solitary Circumscribed Masses

There were no malignancies among the 153 lesions described as multiple bilateral circumscribed benign-appearing masses at US, including among 12 multiple bilateral lesions where the largest lesion was palpable and was described as a complicated cyst (n = 8) or as oval and circumscribed (n = 4). The upper limit of the 95% CI for the observed 0% (zero of 153) malignancy rate was 2.4%; if the 26 cases with less than 2 years of follow-up and no biopsy are excluded, 0% (zero of 127) multiple bilateral circumscribed masses were malignant (95% CI: 0%, 2.9%) (Table 3).

Among 1038 solitary masses with similar descriptions, eight (0.8%; 95% CI: 0.3%, 1.5%) were malignant, including one (0.3%) of 288 complicated cysts, three (0.7%) of 437 oval circumscribed solid masses, two (1.4%) of 140 circumscribed solid masses with two or three gentle lobulations, one (0.9%) of 114 clustered microcysts, and one (1.7%) of 59 round circumscribed solid masses (Table 3). The malignancy rate among oval circumscribed solid masses (0.7% [three of 437]) was not different from that of similar masses with two or three gentle lobulations (1.4% [two of 140]; P = .438). None of these eight benign-appearing malignancies were palpable. At histopathologic examination, two were intermediate-grade ductal carcinomas in situ; two were low-grade, two were intermediate-grade, and one was high-grade invasive ductal carcinoma; and one was an 18-mm node-negative invasive lobular carcinoma. Maximum invasive ductal tumor size was 18 mm (range, 5–18 mm), and only one tumor (a 12-mm mixed invasive and intraductal carcinoma) was node positive. After the exclusion of 202 lesions with less than 2 years of follow-up, three of 836 solitary circumscribed benign-appearing masses were malignant (0.4%; 95% CI: 0.1%, 1.0%). Among 441 solitary masses that looked suspicious at US, 40 (9.1%) were malignant (Table 4).

Table 4.

Malignancy Rates for 1426 Other US-detected Lesions in 2172 Women Screened with US

Note.— Data in parentheses are raw data. These lesions are not described in Table 3.

^*The most common “other” findings were dilated ducts (n = 16), postsurgical collections (n = 11), probable normal variants or fat lobules (n = 8), probable lipomas (n = 8), fat necrosis or oil cysts (n = 6), cysts that were not simple (n = 6), post–vacuum-assisted core biopsy site findings (n = 6), and probable diabetic mastopathy (n = 5).

^†The denominator includes 12 lesions described as multiple bilateral findings (which were not circumscribed masses), none of which were malignant: one postsurgical scar, two lymph nodes, six calcifications without a mass, one intraductal mass, and two “other” lesions.

^‡The denominator includes one benign lesion described as multiple bilateral irregular not-circumscribed masses.

Management of Multiple Bilateral and Solitary Circumscribed Masses

Of 153 multiple bilateral circumscribed masses at US, 107 (69.9%) were considered to be benign, with BI-RADS scores of 1 (n = 1, complicated cyst) or 2 (n = 106, including 79 complicated cysts, four clustered microcysts, and 23 circumscribed oval masses). Another 41 such masses (26.8%) were assessed as BI-RADS category 3 (probably benign) (17 complicated cysts, one cluster of microcysts, 18 circumscribed oval masses, and five circumscribed masses with two or three gentle lobulations), and five such masses (3.3%) were classified as BI-RADS category 4a (one complicated cyst, two circumscribed oval masses, and two circumscribed masses with two or three gentle lobulations). One sonographic BI-RADS category 4a assessment was downgraded to BI-RADS category 2 after integration with mammographic findings; there were no other changes. Six multiple bilateral lesions (3.9%) were recommended for biopsy, and five were sampled for biopsy, after the screening examination when they were first detected, prompted by US findings. Another seven multiple bilateral lesions were sampled for biopsy after the first screening examination because of participant or investigator preference, and three more lesions were sampled for biopsy after a subsequent round of screening; a total of 15 lesions actually were sampled for biopsy.

Among 1038 solitary circumscribed masses, 381 (36.7%) were assessed as BI-RADS category 1 (n = 3, presumably erroneously) or category 2 (benign) (n = 378); 470 (45.3%), as BI-RADS category 3 (probably benign); 171 (16.5%), as BI-RADS category 4a; 14 (1.3%), as BI-RADS category 4b; and one (0.1%) as BI-RADS category 4c at first detection. Of these 1038 masses, 221 (21.3%) were sampled for biopsy when first seen, including two (0.5%) of 378 BI-RADS category 2 masses, 68 (14.5%) of 470 BI-RADS category 3 masses, and 140 (81.9%) of 171 BI-RADS category 4a masses. Another 20 solitary circumscribed masses were sampled for biopsy at a later time, including six BI-RADS category 2 masses, 13 BI-RADS category 3 masses, and one BI-RADS category 4a mass. Of the eight malignant solitary circumscribed masses, five were considered as BI-RADS category 3 masses initially (three of these masses developed suspicious changes of enlargement and not circumscribed margins at 6-, 12-, or 24-month follow-up, and two masses [one ipsilateral and one contralateral] in one participant were diagnosed at mastectomy for concurrent cancer elsewhere), two were classified as BI-RADS category 4a masses, and one was classified as a BI-RADS category 4b mass.

Eighty-two of 135 women (60.7%; 95% CI: 52%, 69%) with multiple bilateral circumscribed masses also had a solitary US-depicted lesion. Malignancies were seen sonographically in two of these 82 women (2.4%; 95% CI: 0.3%, 8.5%) (Figure).

Discussion

Multiple bilateral circumscribed masses are relatively common at US and were seen in 6.2% of participants across three rounds of screening in this study. Follow-up adds cost and can be time consuming if each such mass requires documentation and measurement, particularly with hand-held whole-breast US. It is important that no malignancies were identified among such lesions in this study; however, on the basis of our sample size, the malignancy rate could be as high as 2.9%. Even if there were a few malignancies among such lesions, there was no evidence of adverse patient outcome from diagnostic follow-up at the time of the next annual US examination. As with a BI-RADS category 3 designation (7), such diagnostic follow-up would typically be performed for a period of 2 years, or less if findings decrease or resolve at initial 12-month follow-up. Because multiple bilateral circumscribed masses at screening mammography are considered benign findings (1), we believe that diagnostic follow-up of such masses identified at US is necessary only if there are no correlates at mammography.

Multiple bilateral masses were far more common at screening US than what has been reported at mammography, where 1.7% of consecutive screening examinations in another series (1) showed such findings. In this analysis, we report that, among 2172 participants with two breasts, 135 (6.2%) had multiple bilateral benign-appearing masses at US that were not simple cysts. The lower prevalence at mammography may be caused by masking by dense parenchyma at mammography, although breast density was not detailed in the report of Leung and Sickles (1). Indeed, we found in this series that multiple bilateral circumscribed masses at US increased in prevalence with increasing breast density. In a prior analysis of data from the ACRIN 6666 trial (10), simple cysts were seen in 1255 (47.1%) of all 2662 participants at at least one screening examination over the 3 years, and 602 (22.6%) participants had bilateral simple cysts at at least one of three screening US examinations. In our entire study population, we had only 111 participant cancer events over 2662 women and 7473 screening examinations (ie, 1.5% of examinations were in women given a diagnosis of cancer) (3): Recognition and appropriate management of the much more common cysts and multiple bilateral findings are critical to reducing false-positive findings from screening US.

Of 745 total BI-RADS category 3 lesions in ACRIN 6666 participants (14), 41 (5.5%) were circumscribed benign-appearing multiple bilateral lesions. Another 183 (24.6%) BI-RADS category 3 lesions in ACRIN 6666 were described as complicated cysts, with one of those found to be malignant (14).

We have previously reported (3) that fewer false-positive findings were observed at screening US in women with a personal history of breast cancer than in other women at elevated risk. Relative reduction in false-positive findings in women with a personal history of breast cancer has also been observed with MR imaging screening (15–17). Reduction in both background parenchymal enhancement and cystic change at MR imaging have been noted in the previously irradiated breast (18). It was interesting to note in this series that multiple bilateral findings were also far less common in such women, and we confirmed that cysts were far less common in the previously treated breast. These observations suggest a relative quiescence of the previously affected breast, likely because of radiation therapy, and perhaps of both breasts because of chemoprevention therapies, although the latter were not specifically detailed in this study.

Excluding simple cysts, which are typically benign, fewer than 2% of solitary circumscribed masses at mammography (19–21) or US (22) prove malignant. Across all 2662 participants in ACRIN 6666, clustered microcysts were a common type of circumscribed breast mass, being seen in 104 (3.9%) women (10). No malignancies have been reported across 112 clustered microcysts in the prior literature (23–26); one (0.9%) of the 114 lesions initially described as solitary clustered microcysts in this analysis was described as a complex cystic and solid mass at follow-up, which showed an 18-mm node-negative invasive lobular carcinoma, with no other malignancies. The presence of indistinct margins or a solid component in such lesions should prompt biopsy (10).

Among solitary circumscribed benign-appearing masses in this series, three (0.7%) of 437 oval circumscribed masses and two (1.4%) of 140 masses with two or three gentle lobulations proved to be malignant, as did one (0.3%) of 288 solitary masses thought to be complicated cysts. Across the prior literature (23,25–29), only two (0.2%) of 838 complicated cysts were found to be malignant (one due to a 3-mm ductal carcinoma in situ within a papilloma [28] and one due to a 4-mm grade 2 invasive ductal carcinoma [26]). Complicated cysts and solid masses can be difficult to distinguish, and, indeed, across three series (26–28) (summarized in [10]), 64 (12%) of 517 masses thought to be complicated cysts proved to be solid masses at aspiration, including the two malignancies. Importantly, when either complicated cysts or solid oval circumscribed masses were multiple and bilateral, there were no malignancies in our study. Since solitary circumscribed masses seen at US do appear to merit short-interval follow-up, and multiple bilateral masses may not (assuming our results are further validated in larger studies), performance of bilateral whole-breast US with documentation of masses other than simple cysts at the time of initial US screening may reduce the need for short-interval follow-up if similar bilateral findings are identified.

Although ours was a screening study, we did ask investigators to assess the palpability of findings at the time of the screening US examination. There were no malignancies among the 12 multiple bilateral circumscribed masses for which the largest such mass was palpable. Across five series (30–34), encompassing 1024 palpable probably benign masses at US, only 16 (1.6%) malignancies were identified, and all were node negative. Where detailed (30,32–34), 294 (36%) of 807 such masses were sampled for biopsy, including 51 (7.2%) of 713 masses that grew or developed suspicious changes at follow-up (30,32,33). Those series did not specifically report which lesions, if any, were multiple and bilateral. It would appear that management should derive primarily from imaging findings rather than whether a mass is palpable. Development of any suspicious features or—according to the work of Gordon et al (35)—growth of more than 20% in diameter within 6 months should prompt biopsy (although Leung and Sickles [1] did not observe any malignancies when multiple masses enlarged, among 26 such cases).

This study had several strengths. We considered consecutively accrued participants across a multicenter prospective experience, and follow-up data were available for all included participants. Detailed feature analysis results were prospectively recorded, including the added feature that the lesion was multiple, bilateral, and circumscribed.

There were a few limitations to this analysis. We did not specifically consider the mammographic features of lesions in this analysis, because investigators initially interpreted the US study without benefit of the mammographic study. When both mammographic and US findings were considered to be benign, there was no separate integrated interpretation. For the multiple bilateral lesions assessed as BI-RADS category 3 lesions, integration with mammographic findings did not change this assessment. Investigators included among multiple bilateral lesions six lesions that were described only as calcifications without a mass (but we excluded them for this analysis), even though the intent was to use the multiple bilateral descriptor for bilateral circumscribed oval benign-appearing masses. Clearly, appropriate characterization of calcifications requires mammographic evaluation. For a small subset of 26 (17.0%) of 153 multiple bilateral circumscribed lesions and 202 (19.5%) of 1038 solitary circumscribed benign-appearing findings identified, we had less than 2 years of follow-up data.

In summary, multiple bilateral circumscribed benign-appearing masses are common at screening US, and there were no malignancies (zero of 127 [95% CI: 0%, 2.9%] for lesions with at least 2 years of follow-up) among such masses in this prospective series. As with mammography, it is important to carefully scrutinize the features of each mass at US, as cancer was also present in two (1.7%) of 118 women with multiple bilateral circumscribed masses and at least 2 years of follow-up. We recommend diagnostic follow-up at 1 year for multiple bilateral circumscribed benign-appearing masses identified at screening US, with resumption of screening thereafter if the mass is stable, but we recognize that larger studies need to be performed to document that this is a valid approach.

Advances in Knowledge.

• • Multiple bilateral circumscribed benign-appearing masses are common at bilateral whole-breast US, occurring in 135 (6.2%; 95% confidence interval [CI]: 5.2%, 7.3%) of 2172 women across 3 years of screening in this study.

• • No malignancies were observed among 127 multiple bilateral circumscribed benign-appearing masses detected at US with at least 2 years of follow-up (95% CI: 0%, 2.9%).

• • The malignancy rate among solitary circumscribed masses (complicated cysts; clustered microcysts; circumscribed oval, round, gently lobulated masses) detected at US and with at least 2 years of follow-up was 0.4% (three of 836; 95% CI: 0.1%, 1.0%).

Implications for Patient Care.

• • Recognition that multiple bilateral benign-appearing masses detected at screening US are unlikely to be malignant can potentially reduce unnecessary short-interval follow-up and biopsy.

• • Scrutinizing the features of each mass identified is important, because 82 (60.7%; 95% CI: 52%, 69%) of 135 women with multiple bilateral circumscribed masses also had a solitary lesion at US, and two of the 82 women had cancer (2.4%; 95% CI: 0.3%, 8.5%).

Disclosures of Conflicts of Interest: W.A.B. Financial activities related to the present article: none to disclose. Financial activities not related to the present article: is a member of the medical advisory board of Philips; is a consultant to Naviscan and was a consultant to SuperSonic Imagine; was on the speakers bureau of SuperSonic Imagine; has received payment for manuscript preparation from SuperSonic Imagine and Naviscan; has received payment for development of educational presentations from Gamma Medica; institution has grants or grants pending from Hologic. Other relationships: none to disclose. Z.Z. No relevant conflicts of interest to disclose. J.B.C. No relevant conflicts of interest to disclose. E.B.M. Financial activities related to the present article: none to disclose. Financial activities not related to the present article: is a member of the scientific advisory boards of Hologic, Toshiba Ultrasound, and Quantason; is a member of the board of Seno; is on the clinical advisory committee of Quantason; has been compensated for case reviews by legal firms in Chicago, Ill, and Louisville, Ky; is on the speakers bureaus of Siemens Ultrasound, Toshiba Ultrasound, and SuperSonic Imagine; has received honoraria for lectures from the Institute of Advanced Medical Education, the International Institute for Continuing Medical Education, and MD Anderson Cancer Center; receives royalties from Thieme; has received payment for development of educational presentations from IAME; institution has grants or grants pending from Siemens Ultrasound. Other relationships: none to disclose.