JNCI J Natl Cancer Inst (2016) 108(4): djv367

doi:10.1093/jnci/djv367
First published online December 28, 2015
Article

article
Ultrasound as the Primary Screening Test for Breast
Cancer: Analysis From ACRIN 6666

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

Wendie A. Berg, Andriy I. Bandos, Ellen B. Mendelson, Daniel Lehrer,
Roberta A. Jong, Etta D. Pisano
Affiliations of authors: Magee-Womens Hospital of UPMC (WAB) and Department of Biostatistics (AIB), University of Pittsburgh School of Medicine, Pittsburgh,
PA; Department of Radiology, Feinberg School of Medicine, Northwestern University, Chicago, IL (EBM); CERIM, Buenos Aires, Argentina (DL); Department of
Medical Imaging, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada (RAJ); Department of Radiology, Medical University of South
Carolina, Charleston, SC (EDP).

Correspondence to: Wendie A. Berg, MD, PhD, Department of Radiology, Magee-Womens Hospital, 300 Halket St., Pittsburgh, PA 15213 (e-mail: [email protected]).

Abstract
Background: Mammography is not widely available in all countries, and breast cancer incidence is increasing. We
considered performance characteristics using ultrasound (US) instead of mammography to screen for breast cancer.
Methods: Two thousand eight hundred nine participants were enrolled at 20 sites in the United States, Canada, and
Argentina in American College of Radiology Imaging 6666. Two thousand six hundred sixty-two participants completed
three annual screens (7473 examinations) with US and film-screen (n = 4351) or digital (n = 3122) mammography and had
biopsy or 12-month follow-up. Cancer detection, recall, and positive predictive values were determined. All statistical tests
were two-sided.
Results: One hundred ten women had 111 breast cancer events: 89 (80.2%) invasive cancers, median size 12 mm.
The number of US screens to detect one cancer was 129 (95% bootstrap confidence interval [CI] = 110 to 156), and for
mammography 127 (95% CI = 109 to 152). Cancer detection was comparable for each of US and mammography at 58 of
111 (52.3%) vs 59 of 111 (53.2%, P = .90), with US-detected cancers more likely invasive (53/58, 91.4%, median size 12 mm,

article
range = 2–40 mm), vs mammography at 41 of 59 (69.5%, median size 13 mm, range = 1–55 mm, P < .001). Invasive cancers
detected by US were more frequently node-negative, 34 of 53 (64.2%) vs 18 of 41 (43.9%) by mammography (P = .003).
For 4814 incidence screens (years 2 and 3), US had higher recall and biopsy rates and lower PPV of biopsy (PPV3) than
mammography: The recall rate was 10.7% (n = 515) vs 9.4% (n = 453, P = .03), the biopsy rate was 5.5% (n = 266) vs 2.0%
(n = 97, P < .001), and PPV3 was 11.7% (31/266) vs 38.1% (37/97, P < .001).
Conclusions: Cancer detection rate with US is comparable with mammography, with a greater proportion of invasive and
node-negative cancers among US detections. False positives are more common with US screening.

Worldwide, the number of breast cancer cases is increasing, are increasing worldwide, with 425 000 deaths in 2010, including
with 1.4 million new cases globally in 2008 (1), over 1.6 mil- 68 000 in women age 49  years or younger in developing
lion cases in 2010 (2), and projections of 2.1 million by 2030 (1). countries (2).
Nearly half of this burden is observed in developed countries, While advances in treatment have improved outcomes from
many of which have organized screening. Fully 23% of global breast cancer, axillary lymph node status remains the most
breast cancer cases are seen in women age 15 to 49  years in important prognostic factor. Clinically detected cancers are
developing countries (2). More importantly, even after correcting larger, with median size of 2.6 cm, compared with those found
for the increasing number of cases, deaths from breast cancer with screening mammography, with median size of 1.5 cm (3).

Received: February 11, 2015; Revised: May 27, 2015; Accepted: October 28, 2015
© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: [email protected].

1 of 8
 2 of 8  |  JNCI J Natl Cancer Inst, 2016, Vol. 108, No. 4

Cancers found clinically are more likely to show axillary nodal from all participating sites, ACRIN and National Cancer Institute
metastases: 38% to 45% are node positive compared with 18% to Cancer Imaging Program approval, and data and safety monitor-
25% of cancers detected by mammography screening (4,5). ing committee review every six months.
The mortality benefit from mammographic screening is
because of identification of node-negative invasive cancers (6).
In randomized controlled trials, a 15% reduction in breast can- Screening Methods
cer mortality has been observed in women age 40 to 49  years Each participant underwent three rounds of mammographic and
at entry, and a 22% reduction in women age 50 to 74 years (7). physician-performed ultrasonographic screening examinations
The reduced benefit from mammography in younger women at 0  months (screen 1), 12  months (screen 2), and 24  months
is because of several factors, including the masking effects of (screen 3) in a randomized order assigned prior to initial study
dense parenchyma, which is more common in younger women imaging. Reference standard was available for 2662 unique par-
(8), and also because cancers are more often rapidly growing ticipants: 2659 women screened initially, 2493 women at screen
invasive cancers, which may present clinically in the interval 2, and 2321 women at screen 3. At least two-view mammogra-
between screens (9). phy was performed using either screen-film (n = 4351) or digi-
Mammography is not widely available in developing coun- tal (n  =  3122) technique. Ultrasound was physician performed

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

tries. A test that detects node-negative invasive cancers as well using handheld high-resolution linear array broad bandwidth
as or better than mammography, with cancer detection not lim- transducers with maximum frequency of at least 12 MHz using
ited by breast density, portable, less expensive, and not using standard technique and documentation (13). The radiologist
ionizing radiation, could contribute to breast cancer mortality performing and interpreting the screening US and a different
reduction worldwide. radiologist interpreting the study mammogram were not per-
In the prospective international multicenter American mitted to know the results of the other current screening exami-
College of Radiology Imaging Network (ACRIN) protocol 6666, nation until their interpretations had been recorded, although
a statistically significant increase in cancer detection was prior breast imaging (if any) was available together with risk fac-
observed when physician-performed whole-breast screening tor and biopsy/surgical history.
ultrasound (US) was added to mammography (10,11). In ACRIN Assessments for each breast and lesion were recorded using
6666, screening US was performed and interpreted indepen- Breast Imaging–Reporting and Data System (BI-RADS) (12,14,15).
dently of mammographic results. The study affords the oppor- An expanded seven-point BI-RADS assessment scale was used
tunity to consider performance characteristics of programmatic at the lesion and breast level: 1, negative; 2, benign; 3, probably
breast cancer screening using US alone, while also comparing benign; 4a, low suspicion; 4b, intermediate suspicion; 4c, mod-
results with those observed with screening mammography in erate suspicion; and 5, highly suggestive of malignancy. Further
the same participants. details of screening methods are included in the Supplementary
Materials (available online).
Methods

Participants Reference Standard

Participants were asymptomatic women with heterogeneously Reference standard was cancer based on biopsy within 365 days
or extremely dense breast tissue (12) in at least one quadrant of mammographic screening or no cancer based on a mini-
and at least one other risk factor for breast cancer (prior- mum clinical follow-up of one year. Each mammographic and
itized as in [11] and detailed in the Supplementary Materials, US screen was targeted to occur 365  days after the previous
available online). Participants were at least age 25  years annual screening. A complete examination of all study breasts
article

(median  =  55  years, range  =  25–91  years) at study entry and performed more than 11 full months after the previous screen
provided written, informed consent at their initial visit. Two was considered the next annual screen. A  diagnosis of inva-
thousand eight hundred nine women were recruited from 20 sive or intraductal breast cancer was considered disease posi-
sites (18 in the United States, one in Buenos Aires, Argentina, tive. Further details of reference standard are included in the
and one in Toronto, Ontario, Canada; one other site quali- Supplementary Materials (available online).
fied but did not enroll participants) between April 2004 and
February 2006, of whom 2725 were eligible. Two thousand six
Statistical Considerations
hundred fifty-nine women completed the initial screen with
suitable reference standard of follow-up or biopsy, as did 2493 The primary unit of analysis was the participant (evaluated at
women in year 2 and 2321 women in year 3 (for a total of 7473 an annual screening session). As in the original report of the
paired screening examinations in 2662 unique participants, primary analysis for ACRIN 6666 (10), a participant’s BI-RADs
as detailed in the Supplementary Materials, available online, score and recommendation were derived as the BI-RADS score
and in [11]). Based on self-assigned race/ethnicity, 2467 of 2659 from the breast with cancer, or, for participants without can-
(92.8%) women at first screen were Caucasian, 265 (10%) were cer, the maximum breast-level BI-RADS score. For a participant
Hispanic, 91 (3.4%) were African American, and 90 (3.4%) were with verified cancer diagnosed during the study, the breast with
Asian, with accrual at each site paralleling local population cancer was excluded from analysis for the next annual screen.
demographics. As per the 5th edition of BI-RADS (16), a recommendation for
Web-based data capture and quality monitoring were con- additional diagnostic testing or biopsy prior to the next screen-
ducted by ACRIN’s Biostatistics and Data Management Center ing round (ie, a “recall”) was considered test positive, including
(Center for Statistical Sciences, Brown University, Providence, all BI-RADS assessments of 4a, 4b, 4c, or 5; an assessment of
RI, and ACRIN Headquarters, Philadelphia, PA, respectively). BI-RADS 3 was also considered test positive provided the recom-
The study was Health Insurance Portability and Accountability mendation was for short-interval (usually 6 months) follow-up,
Act–compliant and received institutional review board approval additional imaging, or biopsy.
 W. A. Berg et al.  |  3 of 8

The sensitivity, specificity, recall rate, positive predictive only on US in breasts that were visually more than 80% dense
value of recall (PPV1), and positive predictive values of partic- (Ptrend = .06) (Table 3).
ipant-level biopsies recommended (PPV2) and biopsies per- Despite close similarity in total number of detected cancers,
formed (PPV3) were estimated. Invasive tumor size was recorded. the vast majority of cancers detected by US were invasive (53/58,
Results of nodal staging were reported when performed, but 91.4%, median size = 12 mm, range = 2–40 mm), compared with
nodal staging was not performed in women with a personal his- 41 of 59 (69.5%, median size = 13 mm, range = 1–55 mm) by mam-
tory of axillary nodal dissection prior to study entry. mography (bootstrap P < .001), and invasive cancers depicted by
Data were analyzed using SAS v. 9.3 (SAS Institute Inc., Cary, US were statistically significantly more frequently node nega-
NC). Sensitivity levels and specificity levels estimated for indi- tive (34/53, 64.2%) compared with those seen on mammogra-
vidual years were compared using exact McNemar’s test; 95% phy (18/41, 43.9%, bootstrap P = .003). The differences remained
Wilson confidence intervals (CIs) were provided for individual statistically significant in the cluster bootstrap with sites as re­-
proportions (proc freq, SAS v. 9.3). Comparisons of proportions sampling units. Cancers seen only on screening US were all stage
estimated from the data spanning multiple years (eg, sensitiv- IIA or lower (Table 4). Among 89 invasive cancers, 30 (33.7%) were
ity or specificity for year 2 and 3 screenings combined), as well seen only with US and 18 (20.2%) only with mammography.
as tests for trend, were performed using generalized estimating DCIS was much more likely to be detected by mammogra-

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

equation (GEE) model for binary data (proc genmod, SAS, v. 9.3) phy, with 18 of 22 (81.8%) seen on mammography compared
accounting for possible correlation between assessments of the with only five of 22 (22.7%) by US (P  =  .002, Table  1). Two of 22
same patient. Cluster bootstrap (17) with participants as resa- (9.1%) DCIS were seen only on US (one high nuclear grade and
mpling units, based on 10 000 resamples, was used for estimat- one intermediate nuclear grade).
ing 95% CIs for the difference in rates as well as for individual
proportions estimated from the data combining multiple years.
False Positives
To evaluate the sensitivity of the primary results to clustering
within the 20 centers, we additionally performed cluster boot- For 2659 first study screens with reference standard, of which
strap with sites as resampling units. All presented analyses are over 98% were incidence screens for mammography (ie, a prior
exploratory, following the primary analysis comparing the com- screening mammogram was available) and just over 11% were
bination of mammography and ultrasound to mammography incidence screens for ultrasound (10), US prompted recall of
alone (10,11). The reported P values and 95% confidence inter- more women than mammography (555, 20.9%, vs 306, 11.5%, P
vals are two-sided, with .05 threshold used for statistical signifi- < .001) (Table 1). When comparing incidence screens in years 2
cance assessments. and 3 for both modalities, the recall rate of US at 515 of 4814
(10.7%) was comparable with, although still slightly higher than,
mammography, at 453 of 4814 (9.4%, P  =  .03). Of 7362 screens
Results
in women without cancer, 1012 were recommended for fur-
Cancer Detection ther testing prior to the next screening US (overall specific-
ity = 86.3%). Overall, 810 of 2552 (31.7%, 95% CI = 30.0% to 33.6%)
Across 7473 screens in 2662 unique participants, 110 women unique women without cancer were recalled at least once dur-
were diagnosed with 111 breast cancer events; one participant ing the three screening rounds because of US compared with
was diagnosed initially in year 1, then with contralateral can- 591 of 2552 (23.2%, 95% CI = 21.6% to 24.8%) prompted by mam-
cer in year 3. Of the 111 cancers, 89 (80.2%) were invasive, with mography (P < .001) (Table  5). When results of mammography
median size of 12 mm (range 1 to 55 mm) and 57 of 70 (81%) with and US were integrated, 294 recalls were avoided: 1107 of 2552
nodal staging were node negative. (43.2%) participants without cancer were actually recalled.
The number of US screens needed to detect one cancer At the per-screen level, the overall specificity of US was

article
was 129 (95% CI = 110 to 156), and for mammography 127 (95% lower than that of mammography: 6350 of 7362 (86.3%) vs 6662
CI = 109 to 152). The total number of cancers detected was com- of 7362 (90.5%, P < .001) (Table 1). The difference remained highly
parable across modalities, at 58 of 111 (52.3%) for US and 59 of statistically significant in the cluster bootstrap analysis with
111 (53.2%) for mammography (P = .90, with 95% CI for the differ- sites as resampling units. The age-based and breast density–
ence from -14.4% to 12.6 %) (Table 1). The cancer detection rate based distributions of the false-positive results for individual
with US was 9 per 1000 in year 1 (95% CI = 6.1 to 13.4) and 7.1 per screens are summarized in Table  6. False-positive recalls from
1000 in years 2 and 3 (95% CI = 5.2 to 9.1); these rates were sta- US decreased with increasing patient age (P = .002) (Table 6), a
tistically nonsignificantly different (P = .54 and .53) from those of tendency observed for mammography as well but not statisti-
mammography at 7.5 per 1000 in year 1 (95% CI = 4.9 to 11.6) and cally significant (P = .13). More false positives were seen with US
8.1 per 1000 in years 2 and 3 (95% CI = 6.2 to 10.2). Of 89 invasive with increasing breast density (P = .03) (Table 6), but no consist-
cancers, 53 (59.6%) were detected by US and 41 (46.1%) by mam- ent trend was observed for mammography (P = .25). The greatest
mography (P = .11) (Table 1). increases in false positives with US compared with mammogra-
There were no statistically significant differences in propor- phy (P < .001) were in women age 40 to 69 years and for women
tions of detected cancers in participants of different breast den- with density visually estimated from mammograms at greater
sity or age groups (Table 2). The rate of cancer detection only by than 40%.
US and not mammography appeared to increase with increas- Likelihood of cancer was lower with US than mammography
ing breast density, at three of 17 (17.6%) cancers in breasts that for each of recall (PPV1), biopsies recommended (PPV2), and biop-
were visually 26% to 40% dense mammographically, and six of sies performed (PPV3) (Table 1). For incidence screening (years 2
16 (37.5%) of cancers in breasts that were visually more than and 3), short-term follow-up rates were previously reported (11)
80% dense, though the trend was not statistically significant as 3.9% (190/4814) for US vs 1.6% for mammography (76/4814,
(P  =  .11) and estimate of density was subjective. For invasive and this difference was statistically significant, P < .001); biopsy
cancers, two of 10, 20.0%, were seen only on US in breasts that rate was 5.5% (266/4814) vs 2.0% (97/4814, P < .001), and PPV3 of
were visually 26% to 40% dense and six of 12, 50.0%, were seen biopsies performed was 11.7% (31/266) vs 38% (37/97) (P < .001).
 article

Table 1.  Breast cancer detection by ultrasound or mammography in 2662 participants screened for three annual rounds (7473 screens)*

Mammography Differences US vs US but not

Screening US alone alone mammography mammography
performance
characteristic No./total exams Rate (95% CI†), % No./total exams Rate (95% CI†), % Diff (95% CI†) P‡ No./total exams Rate (95% CI†), %

Cancer detection Rate per 1000 screens

  Year 1 24/2659 9.0 (6.1 to 13.4) 20/2659 7.5 (4.9 to 11.6) 1.5 (-4.8 to 1.9) .54 14/2659 5.3 (3.1 to 8.8)
  Years 2&3 34/4814 7.1 (5.2 to 9.1) 39/4814 8.1 (6.2 to 10.2) -1.0 (-3.6 to 1.5) .53 18/4814 3.7 (2.3 to 5.4)
Sensitivity
 Overall 58/111 52.3 (43.2 to 61.3) 59/111 53.2 (44.1 to 62.2) -0.90 (-14.4 to 12.6) .90 32/111 28.8 (20.7 to 36.9)
  Invasive cancers 53/89 59.6 (49.2 to 69.2) 41/89 46.1 (36.1 to 56.4) 13.5 (-2.2 to 28.1) .11 30/89 33.7 (24.7 to 44.0)
 DCIS 5/22 22.7 (10.1 to 43.4) 18/22 81.8 (61.5 to 92.7) -59.1 (-86.4 to -31.8) .002 2/22 9.1 (2.5 to 27.8)
  Year 1 24/36 66.7 (50.3 to 79.8) 20/36 55.6 (39.6 to 70.5) 11.1 (-15.0 to 36.4) .54 14/36 38.9 (24.8 to 55.1)
  Years 2&3 34/75 45.3 (34.2 to 56.4) 39/75 52.0 (41.0 to 63.0) -6.7 (-23.0 to 9.7) .53 18/75 24.0 (14.6 to 33.8)
Specificity
4 of 8  |  JNCI J Natl Cancer Inst, 2016, Vol. 108, No. 4

 Overall 6350/7362 86.3 (86.1 to 87.8) 6662/7362 90.5 (90.5 to 91.9) -4.2 (-5.3 to -3.2) <.001 552/7362 7.5 (6.9 to 8.2)
  Year 1 2092/2623 79.8 (78.2 to 81.2) 2337/2623 89.1 (87.8 to 90.2) -9.3 (-11.3 to -7.5) <.001 207/2623 7.9 (6.9 to 9.0)
  Years 2&3 4258/4739 89.9 (89.8 to 91.6) 4325/4739 91.3 (91.2 to 92.9) -1.4 (-2.6 to -0.1) .02 345/4739 7.3 (6.5 to 8.1)
Recall rate
  Year 1 555/2659 20.9 (19.4 to 22.5) 306/2659 11.5 (10.4 to 12.8) 9.4 (7.5 to 11.2) <.001 466/2659 17.5 (16.1 to 19.0)
  Years 2&3 515/4814 10.7 (9.8 to 11.6) 453/4814 9.4 (8.6 to 10.3) 1.3 (0.1 to 2.5) .03 430/4814 8.9 (8.1 to 9.8)
PPV1
  Year 1 24/555 4.3 (2.9 to 6.4) 20/306 6.5 (4.3 to 9.9) -2.2 (-4.7 to 0.2) .06 14/466 3.0 (1.8 to 5.0)
  Years 2&3 34/515 6.6 (4.9 to 8.5) 39/453 8.6 (6.6 to 10.8) -2.0 (-4.6 to 0.5) .12 18/430 4.2 (2.5 to 6.0)
PPV2
  Year 1 22/257 8.6 (5.7 to 12.6) 19/83 22.9 (15.2 to 33.0) -14.3 (-22.4 to -6.6) <.001 12/215 5.6 (3.2 to 9.5)
  Years 2&3 33/302 10.9 (8.0 to 14.0) 38/126 30.2 (23.4 to 37.5) -19.2 (-26.5 to -12.4) <.001 17/246 6.9 (4.1 to 9.9)
Biopsy rate, %
  Year 1 233/2659 8.8 (7.7 to 9.9) 65/2659 2.4 (1.9 to 3.1) 6.3 (5.2 to 7.5) <.001 208/2659 7.8 (6.9 to 8.9)
  Years 2&3 266/4814 5.5 (4.9 to 6.2) 97/4814 2.0 (1.7 to 2.4) 3.5 (2.8 to 4.2) <.001 239/4814 5.0 (4.3 to 5.6)
PPV3§
  Year 1 21/233 9.0 (6.0 to 13.4) 19/65 29.2 (19.6 to 41.2) -20.2 (-30.2 to 10.7) <.001 12/208 5.8 (3.3 to 9.8)
  Years 2&3 31/266 11.7 (8.4 to 15.1) 37/97 38.1 (29.7 to 47.3) -26.5 (-35.8 to -17.7) <.001 18/239 7.5 (4.5 to 10.7)

* Some of this information can be found in Table 3 of Berg et al. (11), but this table represents a reanalysis of the data. CI = confidence interval; DCIS = ductal carcinoma in situ; PPV1 = positive predictive value of recall;
PPV2 = positive predictive value of participant-level biopsies recommended; PPV3 = positive predictive value of biopsies actually performed; US = ultrasound.
† Patient-level 95% bootstrap CIs (based on 10 000 resamples) for proportions computed from data combining multiple years. Wilson 95% CIs are reported for simple proportions (proc freq SAS v. 9.3, Cary, NC). (The cluster
bootstrap CIs after accounting for the possible correlation within sites differ only fractionally.)
‡ P value for two-sided McNemar’s test for comparison of simple correlated proportions (proc freq SAS v. 9.3); P value from the two-sided Wald test for the imaging modality (US/mammography) coefficient of generalized
estimating equation models for comparison of proportions over multiple years (proc genmod, SAS v. 9.3).
§ PPV3 = rate of malignancies among biopsies actually performed.

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

 W. A. Berg et al.  |  5 of 8

Table 2.  Breast cancer detection by ultrasound or mammography for categories of visually estimated breast density and participant age at time
of screening across three annual screening rounds

Difference in
sensitivity of US but not
Mammography US vs mammography
Screens with cancer US sensitivity sensitivity mammography detections

Screen No. cancers/ No. detected/ No. detected/ No. detected/

characteristic No. screens (Incidence, %) No. Cancers (%) No. Cancers (%) Estimate P* No. cancers (%)

Density, %
 ≤25 1/128 (0.8) 0/1 (0.0) 0/1 (0.0) 0.0 1.00 0/1 (0.0)
 26–40 17/710 (2.4) 9/17 (52.9) 11/17 (64.7) -11.8 .73 3/17 (17.6)
 41–60 36/2390 (1.5) 18/36 (50.0) 21/36 (58.3) -8.3 .66 9/36 (25.0)
 61–80 41/2890 (1.4) 22/41 (53.7) 18/41 (43.9) 9.8 .54 14/41 (34.1)
 >80 16/1352 (1.2) 9/16 (56.3) 9/16 (56.3) 0.0 1.00 6/16 (37.5)

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

 Ptrend† --- --- --- .65 --- .38 .39 --- --- .11‡
 Unknown 0/3 (0) 0/0 (NA) 0/0 (NA) NA NA 0/0 (NA)
Age, y
 <40 2/289 (0.7) 1/2 (50.0) 2/2 (100) -50.0 1.00 0/2 (0.0)
 40–49 16/1538 (1.0) 8/16 (50.0) 7/16 (43.8) 6.3 1.00 6/16 (37.5)
 50–69 79/4916 (1.6) 39/79 (49.4) 42/79 (53.2) -3.8 .76 20/79 (25.3)
 >69 14/730 (1.9) 10/14 (71.4) 8/14 (57.1) 14.3 .75 6/14 (42.9)
 Ptrend† --- --- --- .27 --- .69 .68 --- --- .68‡

* Two-sided Exact McNemar’s test. NA = not applicable; US = ultrasound.

† Using two-sided Wald test for the factor’s coefficient of the generalized estimating equation model accounting for possible correlation between assessments of the
same patients (proc genmod, SAS, v. 9.3, Cary, NC). The test for trend was performed for the two lowest categories grouped together; conclusions remain the same
with for the test for trend with presented categories.
‡ Care must be taken in interpreting P values for “US but not mammography detections” because of post hoc nature of the analyses and sparse data.

Table 3.  Invasive breast cancer detection by ultrasound or mammography for categories of visually estimated breast density and participant
age at time of screening across three annual screening rounds

US, but not

Difference in US mammography
Screens with cancer US Mammography vs mammography detections

Screen No. cancers/ No. detected/ No. detected/ No. detected/

characteristic No. screens (Incidence, %) No. cancers (%) No. cancers (%) Estimate, % P* No. cancers (%)

Density, %
 ≤25 1/128 (0.8) 0/1 (0.0) 0/1 (0) 0.0 1.00 0/1 (0.0)
 26–40 10/710 (1.4) 6/10 (60.0) 6/10 (60) 0.0 1.00 2/10 (20.0)

article
 41–60 30/2390 (1.3) 16/30 (53.3) 17/30 (56.7) -3.3 1.00 8/30 (26.7)
 61–80 36/2890 (1.2) 22/36 (61.1) 13/36 (36.1) 25.0 0.06 14/36 (38.9)
 >80 12/1352 (0.9) 9/12 (75.0) 5/12 (41.7) 33.3 0.29 6/12 (50.0)
  Ptrend† --- --- .23 --- .19 --- .08 .06‡ ---
 Unknown 0/3 (0) 0/0 (NA) 0/0 (NA) NA 0/0 (NA)
Age, y
 <40 1/289 (0.3) 1/1 (100.0) 1/1 (100.0) 0.0 1.00 0/1 (0.0)
 40–49 14/1538 (0.9) 8/14 (57.1) 5/14 (35.7) 21.4 0.51 6/14 (42.9)
 50–69 61/4916 (1.2) 34/61 (55.7) 28/61 (45.9) 9.8 0.36 18/61 (29.5)
 >69 13/730 (1.8) 10/13 (76.9) 7/13 (53.8) 23.1 0.51 6/13 (46.2)
  Ptrend† --- --- .38 --- .47 --- .94 .80‡ ---

* Two-sided exact McNemar’s test. NA = not applicable; US = ultrasound.

† Using two-sided Wald test for the factor’s coefficient of the generalized estimating equation model accounting for possible correlation between assessments of the
same patients (proc genmod, SAS, v. 9.3, Cary, NC). The test for trend was performed for the two lowest categories grouped together; conclusions remain the same
with for the test for trend with presented categories.
‡ Care must be taken in interpreting P values for “US but not mammography detections” because of the post hoc nature of the analyses and sparse data.

Discussion (approximately $15 000), portable US systems are now equipped

with high-resolution linear transducers (12 MHz or higher) and
Many developing countries lack any screening for breast cancer. are effective at distinguishing simple cysts from suspicious
US is an important test for evaluating palpable breast lumps as masses (20). The equipment used in this study, between 2004
it affords direct correlation of clinical and imaging findings and and 2008, is comparable with what is now available on low-cost
its use has begun in developing countries (18,19). Even low-cost devices. We found that, despite a higher rate of false positives,
 6 of 8  |  JNCI J Natl Cancer Inst, 2016, Vol. 108, No. 4

screening US depicted a similar number of cancers as did mam- malignancy seen only on mammography (22). Mistry et al. (23)
mography but with statistically significantly higher proportions reported that no cancers would have been missed in women age
of invasive and node-negative invasive cancers. 35 to 39  years if United Kingdom best practice guidelines rec-
We are not the first to suggest that US could replace mam- ommending that only US be performed in symptomatic women
mography in some women, though prior reports are limited under age 40 years had been followed. Houssami et al. (24) found
to women with symptoms. In 3129 symptomatic women in higher sensitivity of US than mammography among sympto-
Thailand, US showed an area under the curve of 0.962, which matic women age 45 years and younger at the same specificity.
was better than mammography at 0.954 (P  =  .015), and add- In ACRIN 6666, we found no difference in cancer detection rates
ing mammography to US produced statistically insignificant by US or mammography in categories of age or breast density.
improvement (21). In 1208 focally symptomatic women age 30 One barrier to implementing any screening program is the
to 39  years, Lehman et  al. (22) found higher sensitivity for US harm of false positives. As has been observed in some (25,26), but
than mammography among the 23 (1.9%) women with can- not all (27,28), prior studies of mammography, we found false posi-
cer, with 22 of 23 (95.6%) cancers seen by US and only 14 of 23 tives more common in younger women on US but not mammog-
(60.9%) with mammography (P  =  .0098), albeit with a higher raphy. In our study, with increasing breast density, false positives
false-positive rate for US. They suggest that mammography may increased for US but not mammography, although increasing false

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

have been unnecessary in such women, with only one second positives have been observed in prior studies of mammography
(27,29). Availability of prior comparison examinations reduces
Table 4.  Stage distribution of 111 breast cancer events in 2662 wom- false-positive recalls for all breast imaging modalities to date (30–
en screened with ultrasound and mammography for three years by 32). In this study, recall rates decreased from 20.9% for the first
method of cancer detection screening US to 10.7% in years 2 and 3. Weigert recently reported
(33) that by year 3 technologist-performed screening US across
Both
multiple practices in Connecticut prompted false-positive recalls
US Mammography mammography Clinically
for only 7.7% of women in year 3, compared with 13.8% in year 1
Stage* only only and US MRI† detected‡
(34). In a separate analysis from ACRIN 6666 (35), we showed that
0 2 15 3 1 0 probably benign masses seen only on US, assessed as BI-RADS 3,
I 25 11 9 7 5 can be followed at one year (obviating initial 6-month follow-up or
IIA 5 3 10 1 1 biopsy), which could greatly reduce additional testing prompted
IIB 0 1 1 0 0 by screening US. Similarly, we found no malignancies among mul-
IIIA 0 1 0 0 0 tiple bilateral circumscribed benign-appearing masses identified
IIIB 0 0 2 0 3 on screening US and now recommend BI-RADS 2 assessment (36).
IIIC 0 1 0 0 0 There are several limitations to our analysis. We observed
IV 0 1 1 0 0 more invasive cancers detected by US than by mammography;
however, a larger study is needed to statistically support greater
* According to American Joint Committee on Cancer 7th edition (46). US = ultra-
sensitivity of US to invasive cancers. Only 41.8% (3122/7473) of
sound.
mammograms in this study were performed with digital tech-
† Among a subset of 612 women who had a single screening MRI examination
after the third round of screening mammography and US, nine women were nique, which may slightly underestimate cancer detection by
diagnosed with cancer seen only on MRI, including one woman diagnosed on mammography (25), particularly because our study was enriched
MRI in year 3 after an initial contralateral diagnosis by mammography only in in women with dense breasts. Importantly, we reported no dif-
year 1. ference in supplemental yield of US after digital vs film screen
‡ There were two other cancers detected that were not seen on study imaging
mammography (11). All of our study participants had at least one
or clinically: One woman was diagnosed with ductal carcinoma in situ (stage
risk factor in addition to breast density; cancer detection rates are
article

0) because of computer-assisted detection applied to mammography after

study results were recorded, and another woman with a pathogenic BRCA1 expected to be lower in lower-prevalence populations, and biopsy
mutation was found to have a 7 mm grade 3 invasive ductal carcinoma (stage I) rates may have been artificially high in this population because
on off-study MRI six months after the third screening round. of both patient and radiologist concerns in this elevated-risk

Table 5.  Cumulative unique participants recalled or biopsied because of ultrasound or mammography for 2662 women during the three-year
period

Absolute percent
difference US vs
US Mammography mammography

Performance No./total Rate No./total Rate

characteristic participants (95% CI*) participants (95% CI*) Estimate P†

Overall recall rate 877/2662 32.9 (31.2 to 34.7) 657/2662 24.7 (23.1 to 26.3) 8.26 <.001
Cancer patients recalls 58/110‡ 52.7 (43.5 to 61.8) 59/110 53.6 (44.3 to 62.7) -0.91 1.00
Cancer patients recalls for wrong reason§ 9/110 8.2 (4.4 to 14.8) 7/110 6.4 (3.1 to 12.6) 1.82 .79
Noncancer patients recalls 810/2552 31.7 (30.0 to 33.6) 591/2552 23.2 (21.6 to 24.8) 8.58 <.001
Overall biopsy rate 447/2662 16.8 (15.4 to 18.3) 157/2662 5.9 (5.1 to 6.9) 10.89 <.001
Noncancer patients biopsy (at least one) 390/2552 15.3 (13.9 to 16.7) 100/2552 3.9 (3.2 to 4.74) 11.36 <.001

* 95% Wilson confidence limits for simple proportions. CI = confidence interval; US = ultrasound.
† Two-sided exact McNemar’s test.
‡ One hundred ten women were diagnosed with 111 cancer events (one woman diagnosed in year 1 was diagnosed with contralateral cancer in year 3).
§ Women recalled prior to the appearance of the confirmed cancer or because of finding in a cancer-free location.
 W. A. Berg et al.  |  7 of 8

Table 6.  False positives* by ultrasound or mammography as a function of visually estimated breast density or participant age

Difference in US US, but not mammo­

Screens without cancer US Mammography vs mammography graphy false positives

Screen No. noncancers/ Prevalence No. recalls/ No. recalls/ No. recalls/
characteristic No. Screens (%) No. Noncancers (%) No. Noncancers (%) Estimate (%) P† No. Noncancers (%)

Density, %
 ≤25 127/128 (99.2) 13/127 (10.2) 13/127 (10.2) 0.0 1.00 9/127 (7.1)
 26–40 693/710 (97.6) 84/693 (12.1) 72/693 (10.4) 1.7 0.31 70/693 (10.1)
 41–60 2354/2390 (98.5) 301/2354 (12.8) 200/2354 (8.5) 4.3 <0.001 268/2354 (11.4)
 61–80 2849/2890 (98.6) 422/2849 (14.8) 264/2849 (9.3) 5.5 <0.001 361/2849 (12.7)
 >80 1336/1352 (98.8) 192/1336 (14.4) 151/1336 (11.3) 3.1 0.02 156/1336 (11.7)
 Ptrend‡ --- --- --- .03 --- .25 .16 --- --- .09§
 Unknown 3/3 (100.0) 0/3 (0.0) 0/3 (0.0) 0.0 1.00 0/3 (0.0)
Age, y

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

 <40 287/289 (99.3) 43/287 (15.0) 33/287 (11.5) 3.5 0.28 36/287 (12.5)
 40–49 1522/1538 (99.0) 245/1522 (16.1) 158/1522 (10.4) 5.7 <0.001 203/1522 (13.3)
 50–69 4837/4916 (98.4) 643/4837 (13.3) 443/4837 (9.2) 4.1 <0.001 554/4837 (11.5)
 >69 716/730 (98.1) 81/716 (11.3) 66/716 (9.2) 2.1 0.21 71/716 (9.9)
 Ptrend‡ --- --- --- .002 --- .13 .48 --- --- .02§

* Positive test was defined as Breast Imaging–Reporting and Data System 3 or higher; false positive had no diagnosis of cancer within 365 days of the screening exam.
US = ultrasound.
† Using the two-sided Wald test for the differences between US and mammography in each group (of density or age) within the GEE model accounting for correlation
between examinations of the same patients (proc genmod, SAS v. 9.3).
‡ Using the two-sided Wald test for the factor coefficient of the generalized estimating equation (GEE) model accounting for correlation between examinations of the
same patients (proc genmod, SAS v. 9.3, Cary, NC).
§ Care must be taken in interpreting P values for “US but not mammography false positives” because of the post hoc nature of the analyses.

population. In an average-risk population using an automated alternative to mammography in countries lacking organized
arm for screening US, a 3.6 per 1000 cancer detection rate was screening, particularly with availability of low-cost, portable US
maintained, but only 3% of women were recommended for systems. Where mammography is available, US should be seen
biopsy and 31% of biopsies showed cancer (37). Results from addi- as a supplemental test for women with dense breasts who do
tional screening US series are discussed in the Supplementary not meet high-risk criteria for screening MRI and for high-risk
Materials (available online). Most participants in this study were women with dense breasts who are unable to tolerate MRI (45).
Caucasian, and 94% had breasts less than 4 cm thick (10). High-
frequency US image quality degrades with deep lesions (>3 cm), Funding
and our results would not be generalizable to women with very
large breasts. All ACRIN 6666 radiologist investigators had inter- Funded by The Avon Foundation for Women and the National
preted at least 500 breast US examinations in the preceding two Cancer Institute at the National Institutes of Health, grants
years and successfully completed phantom scanning (38), train- U01CA80098, U01CA79778, and R01CA187593.

article
ing in BI-RADS:US (39), and interpretive skills tasks (40). Using
the same scanning and documentation approach, results with Notes
technologist-performed prevalence screening US to date show
Presented at the 2012 Radiological Society of North America
slightly lower cancer detection rates, PPV1, and PPV3, as summa-
Scientific Assembly.
rized by Berg and Mendelson (13), possibly reflecting lower cancer
Trial registration: Clinicaltrials.gov identifier NCT 00072501.
prevalence in the populations screened. Importantly, Tohno et al.
The study sponsors were not involved in the design of the
(41) reported that technologists in Japan performed better than
study; the collection, analysis, or interpretation of data; the
physicians in detecting cancer during a two-day training course
writing of the manuscript; or decision to submit the manuscript
for handheld US screening. Training would be necessary for any
for publication.
facility planning to offer screening US (42), also true for devel-
We are most grateful to the 100 physician investigators; the
oping countries. With appropriate training, US is no more opera-
many research assistants, ACRIN support staff, Jean Cormack
tor dependent than interpreting mammography (43,44). Finally,
(formerly at the Brown Center for Statistical Sciences), and
while we had previously shown that invasive lobular cancer and
especially to the participants for making this study possible.
low-grade invasive ductal carcinoma are overrepresented among
We also thank Marc Hurlbert (formerly of The Avon Foundation
cancers seen only on US (11), we do not have detailed molecular
for Women) for believing in this work and helping to make it
subtype results for the cancers in this study.
happen. Jeremy Berg deserves special mention at many levels,
In summary, cancer detection by US was shown to be very
including support with statistical analysis.
similar to mammography, and the vast majority of cancers seen
with US were invasive and node negative. While the false-pos-
itive rate of US exceeds that of mammography, the number of References
women recalled for extra testing becomes comparable on inci- 1. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions accord-
dence screening rounds. Although further validation is war- ing to the Human Development Index (2008-2030): a population-based study.
ranted, these results suggest that screening US could be a viable Lancet Oncol. 2012;13:790–801.
 8 of 8  |  JNCI J Natl Cancer Inst, 2016, Vol. 108, No. 4

2. Forouzanfar MH, Foreman KJ, Delossantos AM, et al. Breast and cervical can- raphy in young women with symptoms. AJR Am J Roentgenol. 2003;180(4):935–
cer in 187 countries between 1980 and 2010: a systematic analysis. Lancet. 940.
2011;378(9801):1461–1484. 25. Pisano ED, Gatsonis C, Hendrick E, et  al. Diagnostic performance of digi-
3. Mathis KL, Hoskin TL, Boughey JC, et  al. Palpable presentation of breast tal versus film mammography for breast-cancer screening. N Engl J Med.
cancer persists in the era of screening mammography. J Am Coll Surg. 2005;353(17):1773–1783.
2010;210(3):314–318. 26. Pace LE, Keating NL. A systematic assessment of benefits and risks to guide
4. Vanier A, Leux C, Allioux C, et al. Are prognostic factors more favorable for breast cancer screening decisions. JAMA. 2014;311(13):1327–1335.
breast cancer detected by organized screening than by opportunistic screen- 27. Lehman CD, White E, Peacock S, et  al. Effect of age and breast density on
ing or clinical diagnosis? A study in Loire-Atlantique (France). Cancer Epide- screening mammograms with false-positive findings. AJR Am J Roentgenol.
miol. 2013;37(5):683–687. 1999;173(6):1651–1655.
5. Tabar L, Vitak B, Chen HH, et al. The Swedish Two-County Trial twenty years 28. Johns LE, Moss SM. False-positive results in the randomized controlled trial
later. Updated mortality results and new insights from long-term follow-up. of mammographic screening from age 40 (“Age” trial). Cancer Epidemiol Bio-
Radiol Clin North Am. 2000;38(4):625–651. markers Prev. 2010;19(11):2758–2764.
6. Smith RA, Duffy SW, Gabe R, et  al. The randomized trials of breast cancer 29. Kerlikowske K, Zhu W, Hubbard RA, et al. Outcomes of screening mammog-
screening: what have we learned? Radiol Clin North Am. 2004;42(5):793–806, v. raphy by frequency, breast density, and postmenopausal hormone therapy.
7. Nelson HD, Tyne K, Naik A, et al. Screening for breast cancer: an update for JAMA Intern Med. 2013;173(9):807–816.
the U.S. Preventive Services Task Force. Ann Intern Med. 2009;151(10):727–737, 30. Schell MJ, Yankaskas BC, Ballard-Barbash R, et  al. Evidence-based target

W237-W242. recall rates for screening mammography. Radiology. 2007;243(3):681–689.
8. Kerlikowske K, Ichikawa L, Miglioretti DL, et al. Longitudinal measurement 31. Kriege M, Brekelmans CT, Boetes C, et al. Differences between first and sub-
of clinical mammographic breast density to improve estimation of breast sequent rounds of the MRISC breast cancer screening program for women

Downloaded from https://academic.oup.com/jnci/article/108/4/djv367/2412530 by guest on 16 March 2022

cancer risk. J Natl Cancer Inst. 2007;99(5):386–395. with a familial or genetic predisposition. Cancer. 2006;106(11):2318–2326.
9. Buist DS, Porter PL, Lehman C, et al. Factors contributing to mammography 32. Warner E, Plewes DB, Hill KA, et al. Surveillance of BRCA1 and BRCA2 muta-
failure in women aged 40–49 years. J Natl Cancer Inst. 2004;96(19):1432–1440. tion carriers with magnetic resonance imaging, ultrasound, mammography,
10. Berg WA, Blume JD, Cormack JB, et al. Combined screening with ultrasound and clinical breast examination. JAMA. 2004;292(11):1317–1325.
and mammography vs mammography alone in women at elevated risk of 33. Weigert JM. The Connecticut experiment continues: Ultrasound in the

breast cancer. JAMA. 2008;299(18):2151–2163. screening of women with dense breasts years 3 and 4. In: Radiologic Society of
11. Berg WA, Zhang Z, Lehrer D, et al. Detection of breast cancer with addition of North America. Chicago, IL; 2014.
annual screening ultrasound or a single screening MRI to mammography in 34. Weigert J, Steenbergen S. The Connecticut experiment: the role of

women with elevated breast cancer risk. JAMA. 2012;307(13):1394–1404. ultrasound in the screening of women with dense breasts. Breast J.
12. D’Orsi CJ, Bassett LW, Berg WA, et al. Breast Imaging Reporting and Data System, 2012;18(6):517–522.
BI-RADS: Mammography, 4th edition. Reston: American College of Radiology; 35. Barr RG, Zhang Z, Cormack JB, et  al. Probably Benign Lesions at Screening
2003. Breast US in a Population with Elevated Risk: Prevalence and Rate of Malig-
13. Berg WA, Mendelson EB. Technologist-performed handheld screening breast nancy in the ACRIN 6666 Trial. Radiology. 2013;269(3):701–712.
US imaging: how is it performed and what are the outcomes to date? Radiol- 36. Berg WA, Zhang Z, Cormack JB, et  al. Multiple bilateral circumscribed

ogy. 2014;272(1):12–27. masses at screening breast US: Consider annual follow-up. Radiology.
14. Mendelson EB, Baum JK, Berg WA, et  al. Breast Imaging Reporting and Data 2013;268(3):673–683.
System, BI-RADS: Ultrasound, 1st Ed. Reston: American College of Radiology; 37. Kelly KM, Dean J, Comulada WS, et al. Breast cancer detection using auto-
2003. mated whole breast ultrasound and mammography in radiographically
15. Ikeda DM, Hylton NM, Kuhl CK, et al. Breast Imaging Reporting and Data System, dense breasts. Eur Radiol. 2010;20(3):734–742.
BI-RADS: Magnetic Resonance Imaging. Reston: American College of Radiology; 38. Berg WA, Blume JD, Cormack JB, et al. Lesion detection and characterization
2003. in a breast US phantom: results of the ACRIN 6666 Investigators. Radiology.
16. Sickles EA, D’Orsi CJ. Follow-up and outcome monitoring. In: ACR BI-RADS 2006;239(3):693–702.
Atlas, Breast Imaging Reporting and Data System. Reston, VA: American College 39. Mendelson EB, Böhm-Vélez M, Berg WA, et  al. ACR BI-RADS Ultrasound.

of Radiology; 2013. In. ACR BI-RADS Atlas, Breast Imaging Reporting and Data System. Reston, VA:
17. Field CA, Welch AH. Bootstrapping clustered data. JR Statist Soc B.
American College of Radiology; 2013.
2007;69(3):369–390. 40. Berg WA, Blume JD, Cormack JB, et  al. Training the ACRIN 6666 Investiga-
18. Gonzaga MA. How accurate is ultrasound in evaluating palpable breast
tors and Effects of Feedback on Breast Ultrasound Interpretive Performance
masses? Pan Afr Med J. 2010;7:1. and Agreement in BI-RADS Ultrasound Feature Analysis. AJR Am J Roentgenol.
19. Irurhe NK, Adekola OO, Awosanya GO, et al. The accuracy of ultrasonography 2012;199(1):224–235.
in the diagnosis of breast pathology in symptomatic women. Nig Q J Hosp 41. Tohno E, Takahashi H, Tamada T, et  al. Educational program and testing
Med. 2012;22(4):236–239. using images for the standardization of breast cancer screening by ultra-
20. Hilton SV, Leopold GR, Olson LK, et al. Real-time breast sonography: applica- sonography. Breast Cancer. 2012;19(2):138–146.
tion in 300 consecutive patients. AJR Am J Roentgenol. 1986;147(3):479–486. 42. Mendelson EB, Berg WA. Training and standards for performance, interpreta-
21. Chairat R, Puttisri A, Pamarapa A, et al. Are both ultrasonography and mam- tion, and structured reporting for supplemental breast cancer screening. AJR
article

mography necessary for cancer investigation of breast lumps in resource- Am J Roentgenol. 2015;204(2):265–268.
limited countries? ISRN Oncol. 2013;2013:257942. 43. Berg WA, Blume JD, Cormack JB, et  al. Operator dependence of physician-
22. Lehman CD, Lee CI, Loving VA, et al. Accuracy and value of breast ultrasound performed whole-breast US: lesion detection and characterization. Radiol-
for primary imaging evaluation of symptomatic women 30–39 years of age. ogy. 2006;241(2):355–365.
AJR Am J Roentgenol. 2012;199(5):1169–1177. 44. Bosch AM, Kessels AG, Beets GL, et al. Interexamination variation of whole
23. Mistry SG, Barnes N, Ooi J. Will adherence to new guidance lead to missed breast ultrasound. Br J Radiol. 2003;76(905):328–331.
cancer diagnoses? Evaluation of limiting symptomatic mammograms to 45. Berg WA. Tailored supplemental screening for breast cancer: what now and
over forties. Breast J. 2013;19(2):142–148. what next? AJR Am J Roentgenol. 2009;192(2):390–399.
24. Houssami N, Irwig L, Simpson JM, et  al. Sydney Breast Imaging Accuracy 46. Edge SB, Byrd DR, Compton CC, et al. AJCC Cancer Staging Handbook, 7th ed.
Study: Comparative sensitivity and specificity of mammography and sonog- New York: Springer; 2011.