women with elevated breast cancer risk nih public access

Detection of Breast Cancer with Addition of Annual ScreeningUltrasound or a Single Screening MRI to Mammography inWomen with Elevated Breast Cancer Risk

Wendie A. Berg, MD, PhD1,2, Zheng Zhang, PhD3, Daniel Lehrer, MD4, Roberta A. Jong,MD5, Etta D. Pisano, MD6,7, Richard G. Barr, MD, PhD8, Marcela Böhm-Vélez, MD9, Mary C.Mahoney, MD10, W. Phil Evans III, MD11, Linda H. Larsen, MD12, Marilyn J. Morton, DO13,Ellen B. Mendelson, MD14, Dione M. Farria, MD15, Jean B. Cormack, PhD3, Helga S.Marques, MS3, Amanda Adams, MPH3, Nolin M. Yeh, MS3, Glenna Gabrielli, BS16, and forthe ACRIN 6666 Investigators1American College of Radiology Imaging Network, Lutherville, MD3Center for Statistical Sciences, Brown University, Providence, RI4CERIM, Buenos Aires5Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario6University of North Carolina, Chapel Hill, NC8Radiology Consultants/Forum Health, Youngstown, OH9Weinstein Imaging Associates, Pittsburgh, PA10University of Cincinnati Medical Center, Cincinnati, OH11University of Texas Southwestern Medical Center, Dallas, TX12Keck School of Medicine, University of Southern California, Los Angeles, CA13Mayo Clinic, Rochester, MN14Feinberg School of Medicine, Northwestern University, Chicago, IL15Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO16American College of Radiology, Philadelphia, PA

AbstractContext—Annual ultrasound (US) may detect small, node-negative breast cancers not seen onmammography (M). MRI may depict additional breast cancers beyond mammography andultrasound (M+US).

Objective—Determine supplemental cancer detection yield of ultrasound and MRI in women atelevated risk for breast cancer.

Design, Setting, Participants—From April 2004 to February 2006, 2809 women at 21 siteswith elevated cancer risk and dense breasts consented to three annual independent screens withmammography and ultrasound in randomized order. After 3 rounds of mammography and USscreening, 703 women from 14 sites consented to a single MRI.

Main Outcome Measures—Cancer detection rate (yield), sensitivity, specificity, positivepredictive value of biopsies performed (PPV3 – rate of malignancy among cases positive onscreening, who actually underwent biopsy), interval cancer rate. The diagnosis of breast cancer

NIH Public AccessAuthor ManuscriptJAMA. Author manuscript; available in PMC 2014 January 15.

Published in final edited form as:JAMA. 2012 April 4; 307(13): 1394–1404. doi:10.1001/jama.2012.388.

NIH

-PA Author Manuscript

NIH


NIH


was based on a biopsy showing in situ or infiltrating ductal carcinoma or infiltrating lobularAddress correspondence to WAB: Department of Radiology, Magee-Womens Hospital, 300 Halket St., Pittsburgh, PA 15213,[email protected], 412-641-1635.2Present address: Magee-Womens Hospital, University of Pittsburgh School of Medicine, Pittsburgh, PA7Present address: Department of Radiology, Medical University of South Carolina, Charleston, SC

Presented at the 2009 Radiologic Society of North America Scientific Assembly Trial Registration Clinicaltrials.gov identifier: NCT00072501.

The full protocol is available online at: http://acrin.org/Portals/0/Protocols/6666/Protocol-ACRIN%206666%20Admin%20Update%2011.30.07.pdf

Author Contributions: Drs. Zhang and Cormack had full access to all of the data in the study and take responsibility for the integrityof the data and the accuracy of the data analysis.Study concept and design: Berg, Zhang, Mendelson, Pisano.Acquisition of data: Berg, Mendelson, Lehrer, Böhm-Vélez, Pisano, Jong, Evans, Morton, Mahoney, Hovanessian-Larsen, Barr,Farria, Gabrielli.Analysis and interpretation of data: Berg, Zhang, Cormack, Pisano, Marques, Adams, Yeh.Drafting of the manuscript: Berg, Zhang.Critical revision of the manuscript for important intellectual content: Berg, Zhang, Mendelson, Lehrer, Böhm-Vélez, Pisano, Jong,Evans, Morton, Mahoney, Hovanessian-Larsen, Barr, Farria, Gabrielli.Statistical analysis: Zhang, Cormack, Marques, Adams, Yeh.Obtained funding: Berg.Administrative, technical, or material support: Berg, Mendelson, Gabrielli.Study supervision: Berg, Zhang.

Financial Disclosures: Dr Berg reports that she has served as a consultant to Naviscan, Inc. and SuperSonic Imagine, has receivedresearch support from Naviscan Inc., has prepared educational materials for Gamma Medica, has a research grant from Hologic, Inc.,and is on the medical advisory board of Philips. Dr Mendelson reports that she is a member of the scientific advisory boards ofMediPattern, Hologic, and Siemens and has received equipment support from Philips and research support from SuperSonic Imagineand Siemens. Dr Böhm-Vélez reports that she is a member of the scientific advisory board of Philips, does clinical validation studiesfor Philips Ultrasound, and is on the speakers bureau of Dilon. Dr Pisano reports that her laboratory received research support fromGE Healthcare, Konica Minolta, Sectra AB, Naviscan Inc., Koning, Zumatek, Inc., equipment grants from R2 and iCAD, is a boardmember of ACR Imaging Metrix and NextRay, Inc., and a stockholder inNextRay, Inc. Dr Jong reports that she is a consultant to andreceives research support from GE Healthcare. Dr Evans reports that he is a member of the scientific advisory board of Hologic. DrMahoney reports that she is a consultant to Ethicon EndoSurgery and SenoRx and on the scientific advisory board of Hologic andreceives research support from Naviscan, Inc.. Dr Larsen reports that she receives equipment support from Naviscan Inc. Dr Barrreports that he is a member of the ultrasound advisory boards of and has received equipment support, research support, and speakersfees from Siemens and Philips, an equipment grant from SuperSonic, Inc., and a research grant from Bracco. The remaining coauthorsreport no financial disclosures.

ACRIN 6666 Site Investigators: Allegheny-Singer Research Institute, Pittsburgh, Pennsylvania: William R. Poller, MD, principalinvestigator (PI), Michelle Huerbin, research associate (RA); American Radiology Services–Johns Hopkins Green Spring, Baltimore,Maryland: Wendie A. Berg, MD, PhD (PI), Barbara E. Levit, RT(RA), Kathy Wetzel (RA); Beth Israel Deaconess Medical Center,Boston, Massachusetts: Janet K. Baum, MD, and Valerie J. Fein-Zachary, MD (PIs), Suzette M. Kelleher, BA (RA); CERIM, BuenosAires: Daniel E. Lehrer, MD(PI), Maria S. Ostertag (RA); Duke University Medical Center, Durham, North Carolina: Mary ScottSoo, MD (PI), Brenda N. Prince, RT (RA); Mayo Clinic, Rochester, Minnesota: Marilyn J. Morton, DO (PI), Lori M.Johnson, AAS (RA); Feinberg School of Medicine, Northwestern University, Chicago, Illinois: Ellen B. Mendelson, MD (PI),Marysia Kalata, AA (RA); Radiology Associates of Atlanta, Atlanta, Georgia: Handel Reynolds,MD(PI), Y. Suzette Wheeler, RN,MSHA (RA); Radiology Consultants/Forum Health, Youngstown, Ohio: Richard G. Barr, MD, PhD (PI), Marilyn J. Mangino, RN(RA); Radiology Imaging Associates, Denver, Colorado: A. Thomas Stavros,MD(PI), Margo Valdez (RA); Sunnybrook HealthSciences Centre, University of Toronto, Toronto, Ontario, Canada: Roberta A. Jong,MD(PI), Julie H. Lee, BSC (RA); ThomasJefferson University Hospital, Philadelphia, Pennsylvania: Catherine W. Piccoli, MD, and Christopher R.B. Merritt, MS,MD(PIs),Colleen Dascenzo (RA); David Geffen School of Medicine at University of California Los Angeles Medical Center, Los Angeles:Anne C. Hoyt, MD (PI), Roslynn Marzan, BS (RA); University of Cincinnati Medical Center, Cincinnati, Ohio: Mary C. Mahoney,MD (PI), Monene M. Kamm, AS (RA); University of North Carolina, Chapel Hill: Etta D. Pisano, MD (PI), Laura A. Tuttle, MA(RA), Keck School of Medicine, University of Southern California, Los Angeles: Linda H. Larsen, MD(PI), Christina E. Kiss, AA(RA); University of Texas M. D. Anderson Cancer Center, Houston: Gary J. Whitman, MD (PI), Sharon R. Rice, AA (RA); Universityof Texas Southwestern Medical Center, Dallas: W. Phil Evans, MD (PI), Kimberly T. Taylor, AA (RA); Washington UniversitySchool of Medicine, St. Louis, Missouri: Dione M. Farria, MD, MPH (PI), Darlene J. Bird, RT, AS (RA); and Weinstein ImagingAssociates, Pittsburgh, Pennsylvania: Marcela Böhm-Vélez, MD, (PI), Antoinette Cockroft (RA).

Additional Contributions: The authors thank Jeffrey Blume, PhD, Vanderbilt University, Nashville, TN, for his contributions toinitial study design and supervision and Robert A. Smith, PhD, American Cancer Society, Atlanta, GA, for review and helpfuldiscussions. We especially thank Marydale DeBor, JD, Chief Advisor Avon Breast Cancer Crusade 1993–2003, who was instrumentalin securing the support within the Avon Foundation which made this study possible, and Marc Hurlbert, PhD, of the Avon Foundationfor his continued vision and unwavering support. We are indebted to the many investigators, coinvestigators, and research associatesat the clinical sites. We appreciate the efforts of ACRIN Data Management and Imaging staff, and we especially thank Cynthia Olson,MBA, MHS, for administrative oversight. We thank Eric Berns, PhD, University of Colorado, Denver, CO, for ultrasound imagequality control and R. Edward Hendrick, PhD, University of Colorado, Denver, CO, for MRI imaging quality control. No one wascompensated beyond their usual salary for their efforts for this study. We are especially grateful to the 2809 women who enrolled inthis study.

Berg et al.

JAMA. Author manuscript; available in PMC 2014 January 15.

NIH


NIH


NIH


http://Clinicaltrials.gov

http://acrin.org/Portals/0/Protocols/6666/Protocol-ACRIN%206666%20Admin%20Update%2011.30.07.pdf

http://acrin.org/Portals/0/Protocols/6666/Protocol-ACRIN%206666%20Admin%20Update%2011.30.07.pdf

carcinoma in the breast or axillary lymph nodes. Reference standard was defined as a combinationof pathology and 12-month follow-up and was available for 2662 women (7473 M+US screens)and 612 MRI participants.

Results—The 2662 patients underwent 7473 mammograms and US, with 110 women having111 breast cancers detected, of which 33 were detected on mammography only, 32 on US only, 26on both mammography and US, and 9 on MRI after mammography and US. Eleven were notdetected by any imaging modality. Supplemental incidence-screening US identified 3.7 cancersper 1000 women-screens (95% CI 2.1 to 5.8, p<.001). Sensitivity, specificity, and PPV3 for M+US were 57/75 (0.76, 95% CI 0.65 to 0.85), 3987/4739 (0.84, 95% CI 0.83 to 0.85), and 55/339(0.16, 95% CI 0.12 to 0.21); and for mammography alone 39/75 (0.52, 95% CI 0.40 to 0.64),4325/4739 (0.91,95% 0.90 to 0.92), and 37/97 (0.38, 95% CI 0.28 to 0.49) (p<.001 allcomparisons). Of 612 analyzable MRI participants, 16 (2.6%) had breast cancer diagnosed.Supplemental yield of MRI was 14.7 per 1000 (95% CI 3.5 to 25.9, p=.004). Sensitivity,specificity, and PPV3 for MRI+M+US were 16/16 (1.00, 95% CI 0.79 to 1.00), 390/596 (0.65,95% CI 0.61 to 0.69), and 15/81 (0.19, 95% CI 0.11 to 0.29); and for M+US 7/16 (0.44, 95% CI0.20 to 0.70, p=.004), 503/596 (0.84, 95% CI 0.81 to 0.87, p <.001), and 7/38 (0.18, 95% CI 0.08to 0.34, p= .98) for M+US. Number of screens needed to detect one cancer was 127(95%CI 99 to167) for mammography; 234(95%CI 173 to 345) for supplemental ultrasound, and 68 (95%CI 39to 286) for MRI after negative M+US.

Conclusions—The addition of screening ultrasound or MRI to mammography in women atincreased risk of breast cancer resulted in a higher cancer detection yield but also an increase infalse positive findings.

INTRODUCTIONSix previous single-center studies1–6 and three multicenter trials7–9 have shownsupplemental screening breast ultrasound significantly increases detection of node-negativeinvasive breast cancer in women with mammographically dense breast tissue on the first,prevalence screen, consistently increasing cancer detection (yield) by 3.5 per 1000 in single-center studies and 4.2 to 4.4 per 1000 in multicenter trials. The vast majority of cancers seenonly on ultrasound have been node-negative invasive breast cancers. Until now, it wasunclear whether or not there was a detection benefit to continuing ultrasound screeningannually (i.e. incidence screening).

A substantial majority of ACRIN 6666 participants were at intermediate risk for breastcancer, with over half having personal history of breast cancer (PHBC).7 While there wasevidence from prior studies10–13 that MRI provided considerable detection benefit even aftercombined ultrasound and mammography in high-risk women, the combination of ultrasoundand mammography might still identify the vast majority of cancers when they are nodenegative, at much lower cost to the health care system than MRI, particularly in screeningwomen with a lower prevalence of disease. A substudy of ACRIN 6666 participants wastherefore undertaken to assess the rate and stage of cancers detected with a single screeningMRI.

METHODSStudy Design

Participants were asymptomatic women with heterogeneously dense or extremely densebreast tissue14 and at least one other risk factor for breast cancer (Table 1) who presented forroutine annual mammography. Race and ethnic group were self-assigned based on pre-assigned fixed categories.

Berg et al.


NIH


NIH


NIH


Each participant underwent mammographic and physician-performed ultrasonographicscreening examinations in randomized order, with the interpreting radiologist for eachexamination masked to results of the other study, at 0 months (screen 1), 12 months (screen2), and 24 months (screen 3). The randomization process has been previously described,7

and initial randomization order was maintained for subsequent screening rounds. Ifrecommendation from either screening test was other than routine annual screening, the testwas considered positive and a qualified site investigator then recorded an integratedinterpretation by reviewing study mammogram and ultrasound together, and clinicalmanagement was based on integrated interpretation. If both modalities recommended routineannual follow-up, no integration was performed. Cancers positive only on a given modalityrefers to those not visible on any other modality. Sensitivity of a modality alone refers to thenumber of cancers that would have been detected if only that modality had been used, andincludes some cancers that were also visible on the other modality.

To be eligible for the MRI substudy, women had to have completed the third round ofannual screening ultrasound and mammography per protocol7 and undergo contrast-enhanced breast MRI within 8 weeks of the 24 month (third) screening mammogram.Interpretation of each of screening mammography, screening US, and MRI was blinded toresults of the other examinations. A separate integrated breast-level interpretation across allmodalities (mammography, ultrasound, MRI) was then performed which determined clinicalmanagement. We have previously reported on the slightly greater risk and younger age ofwomen accepting vs. declining MRI.15 Women enrolled at sites in the MRI substudy wereless likely to have PHBC; no other systematic differences were noted across sites.

Web-based data capture and quality monitoring were conducted by the ACRIN biostatisticsand data management center. The study was HIPAA-compliant and received institutionalreview board approval from all participating sites, ACRIN and National Cancer InstituteCancer Imaging Program approval, and data and safety monitoring committee review everysix months.

ParticipantsAmong the 21 sites, 2809 women were recruited between April 2004 and February 2006, ofwhom 2725 were eligible (Figure 1). Women aged at least 25 years presenting for routinemammography were eligible to participate if they met study definitions of elevated risk(Table 1) and had heterogeneously dense or extremely dense parenchyma14 in at least onequadrant, either by prior mammography report or review of prior mammograms. Womenwere excluded if they were pregnant or lactating, had known metastatic disease, signs orsymptoms of breast disease, breast surgery within prior 12 months, or breast implants. Forthe MRI substudy, women also could not have contraindications to MRI [pacemaker,aneurysm clip, other metallic implant, weight > 300 lb, or renal impairment (glomerularfiltration rate of <30 mL/min/1.73 m2 or dialysis regimen)]. Participants provided writteninformed consent at their initial visit and again for the MRI substudy (when applicable).

Screening methods are detailed in the online supplement. Expanded 7-point BI-RADS27,29,30 assessment scale was used: 1, negative; 2, benign; 3, probably benign; 4a, lowsuspicion; 4b, intermediate suspicion; 4c, moderate suspicion; and 5, highly suggestive ofmalignancy.

Reference StandardReference standard was the most severe of biopsy results within 365 days of mammographicscreening and/or clinical follow-up at one year. Each mammographic screen was targeted365 days after the previous mammographic screening. A complete examination of all study

Berg et al.


NIH


NIH


NIH


breasts performed more than 11 full months after the previous screen was considered thenext annual screen; only 88/7473 (1.2%) visits occurred before 11 months. The absence of aknown diagnosis of cancer on participant interview, review of medical records, or both, atleast 11 full months (330 days) after mammographic screening, was considered diseasenegative, as were 7 cases of prophylactic mastectomies with no evidence of cancer atpathology. Biopsy results showing breast cancer (in situ or infiltrating ductal carcinoma orinfiltrating lobular carcinoma) in the breast or axillary lymph nodes were considered diseasepositive.

Statistical MethodsThe primary unit of analysis was the participant. Participant’s BI-RADS score was derivedas the maximum breast level BI-RADS or the score from the cancer side when only onebreast had cancer. In keeping with planned revisions to BI-RADS (Edward A. Sickles, MD,Professor of Radiology, University of California, San Francisco, personal writtencommunication 11/29/09), a screening BI-RADS assessment score of 3, 4a, 4b, 4c, or 5 wasconsidered test positive, provided the recommendation was for other than routine screening.This differs from the definition of test positive used in our initial publication of the firstscreen wherein an assessment of 4a or higher was considered test positive29: results of thefirst screen have been reanalyzed and included herein. For a participant diagnosed withcancer, the breast(s) with cancer were excluded from analysis for the next annual screen.

The cancer detection rate (yield, i.e., the proportion of women with a positive screen test andpositive reference standard), sensitivity, specificity, recall rate, PPV1 (malignancy rateamong cases positive on screening), short-term follow-up rate, biopsy rate and AUC (areaunder the empirical ROC curve using BI-RADS scores) were reported. PPV3 is defined asrate of malignancy among cases positive on screening who underwent biopsy of the samelesion.14 Interval cancers were defined as those diagnosed because of a clinical abnormalitysuch as a lump, skin thickening, or pathologic nipple discharge, in the interval betweenprescribed screenings (i.e. less than 365 days after the last screening mammogram).

Single-year estimates of yield, sensitivity, specificity, recall rate, PPV1, short-term follow-up rate, biopsy rate and PPV3, were determined as simple proportions with exact 95% CIs(Clopper-Pearson).16 The 95% CIs for differences in yield, sensitivity, specificity, recallrate, short-term follow-up rate and biopsy rate were calculated per Fleiss et al17 and p-valuesfor the above comparisons were based on McNemar’s test statistic. The 95% CIs and p-values for differences in PPV1 and PPV3 were calculated using bootstrap resamplingmethod18. All inferences for incidence screens were based on bootstrap resampling method.Estimates, 95% CIs and P-values related to AUC were derived by using method of Delonget al19 for empirical ROC curves. Results in participants with PHBC were compared to thosewithout by the bootstrap method. All p-values were reported as two-sided, with 0.05 set asthreshold for significance. All analyses were implemented in SAS 9.2 (Cary, NC).

RESULTSParticipant Demographic Information

There were 2659, 2493, and 2321 eligible women with reference standard for the first,second, and third annual mammogram and ultrasound screens respectively (Figure 1, Table1). Participant demographics at enrollment were previously reported.7 Median age atenrollment was 55 years (range 25–91). Approximately 29% of women were under age 50 atenrollment and 23% were premenopausal (Table 1). Nearly 54% of women had PHBC.Median age of 612 MRI participants was 57 years (range 27–87); 21% were under age 50 atscan, 25% were premenopausal, and 45% had PHBC.

Berg et al.


NIH


NIH


NIH


Cancer DetectionA total of 110 participants were diagnosed with breast cancer over three years of study, withone woman diagnosed by mammography in year one again diagnosed in the contralateralbreast in year 3 (by MRI only) and counted as separate events, i.e. total of 111 participant-cancer events. Of 111 diagnoses, 89 (80%) were invasive (Table 2). Fifty-nine of 111 (53%)were seen on mammography, including 33 (30%) seen only on mammography; 32 (29%)additional only on US; 9 (8%) only on MRI after both mammography and US; and 11 (10%)not seen on imaging. Of 32 cancers seen only on US, 30 (94%) were invasive, with mediansize 10 mm (range 2–40), and 26/27 (96%) of those staged were node negative.

A total of 16/612 (2.6%) MRI participants were diagnosed with breast cancer, with 12/16(75%) invasive. 9/16 (56%) were seen only on MRI after negative mammography and US:8/9 (89%) were invasive, with median size 8.5 mm (range 1–25), and all 7 with staging werenode negative (Table 2). Two invasive cancers detected by US and not mammography in theMRI substudy were also seen on MRI.

Supplemental Cancer Detection YieldSupplemental US increased cancer detection with each annual screen beyond that ofmammography, adding detection of 5.3 cancers per 1000 women in year one (95%CI 2.1 to8.4; p<.001 vs. mammography alone); 3.7 per 1000 per year in years two and three (95%CI2.1 to 5.8, p<.001 vs. mammography alone) (Table 3); and averaging 4.3 per 1000 for eachof three rounds of annual screening. Addition of MRI increased cancer detection, withsupplemental cancer detection yield of 14.7 per 1000 women (95%CI 3.5 to 25.9, p=.004 vs.M+US) (Table 4). Number of screens needed to detect one cancer was 127(95%CI 99 to167) for mammography; 234(95%CI 173 to 345) for supplemental ultrasound and 68(95%CI 39 to 286) for MRI after negative M+US.

Sensitivity, Specificity, and AUCSensitivity of combined mammography and US was 57/75(0.76, 95%CI 0.65 to 0.85) forcombined screens two and three, higher than mammography alone [39/75(0.52, 95%CI 0.40to 0.64), p<.001]. Specificity of combined mammography and US was 3987/4739 (0.84,95%CI 0.83 to 0.85) for incidence screens, lower than specificity of mammography alone[4325/4739 (0.91, 95%CI 0.90 to 0.92), p<.001] (Table 3).

For 612 MRI participants, sensitivity increased from 7/16(0.44, 95%CI 0.20 to 0.70) withcombined mammography and ultrasound to 16/16(1.00, 95%CI 0.79 to 1.00) with additionof MRI(p=.004). Specificity was reduced to 390/596(0.65, 95%CI 0.61 to 0.69) after MRI,compared to combined M+US at 503/596(0.84, 95%CI 0.81 to 0.87, p <.001) (Table 4).

Overall AUC increased in each year when US was added to mammography (Table 3).Adding MRI lowered apparent performance of M+US since more cancers were identified byMRI (Table 4).

Additional Biopsies and PPV3PPV3 of combined mammography and US was 31/272(0.11, 95%CI 0.08 to 0.16) for year 1and 55/339(0.16, 95%CI 0.12 to 0.21) for incidence screens (years two and three combined),These values were significantly lower than those of mammography alone (19/65 [0.29,95%CI 0.19 to 0.42] and 37/97[0.38, 95%CI 0.28 to 0.49] respectively, p<.001 for both)(Table 3). Percentage of women undergoing biopsy after mammography and US decreasedfrom 272/2659 (10.2%, 95%CI 9.1 to 11.4%) in year 1 to 339/4814 (7.0%, 95%CI 6.3 to7.8%) for incidence screens (p<.001). The biopsy rates after mammography alone were

Berg et al.


NIH


NIH


NIH


65/2659 (2.4%, 95%CI 1.9 to 3.1%) in year 1 and 97/4814 (2.0%, 95%CI 1.6 to 2.5%) forincidence screens.

For 612 MRI participants, rate of biopsy after full workup of M+US was 38/612(6.2%,95%CI 4.4 to 8.4%), which increased to 81/612(13.2%, 95% CI10.7 to 16.2%) with additionof MRI (p<.001). PPV3 after M+US was 7/38(0.18, 95%CI 0.08 to 0.34) and with additionof MRI was 15/81(0.19, 95%CI 0.11 to 0.29, p=.98) (Table 4).

Interval CancersOf 20 women with cancer not seen on either mammography or US in three annual rounds, 9were among the 612 participants who also had MRI screening in year three, with cancerdetected on MRI. Another 9 cancers were identified because of clinical abnormalities ininterval between screens (interval cancer rate 8.1%): two had clinical findings in year one;four in year two; and three in year three. One participant was found to have high-gradeDCIS because of off-study computer-assisted detection applied to mammogram (revealingcalcifications) after year 3 interpretation had been recorded. One participant with BRCA-1mutation had MRI off study six months after the third screen and was found to have a 7 mmnode-negative grade III IDC.

Women with Personal History of Breast Cancer (PHBC)A total of 1426/2659(54%) of participants had PHBC at study entry and underwent 4010screens; 59/1426 (4.1%) were diagnosed with cancer (28 only ipsilateral and 29 onlycontralateral to the original cancer; 2 bilateral). Supplemental yield of US was the same inwomen with PHBC and in women without PHBC (online Table 5A), as was the absoluteincrease in sensitivity due to added ultrasound. Supplemental US was less likely to promptunnecessary recall or biopsy in women with PHBC than those without (online Table 5A). .The supplemental yield of MRI in women with or without PHBC in MR substudy is detailedonline (online Table 5B). Supplemental MRI was less likely to prompt unnecessary recall orbiopsy in women with PHBC than those without (online Table 5B).

DISCUSSIONIn this study, annual supplemental incidence screening ultrasound detected an additional 3.7cancers per 1000 women per year screened beyond mammography alone. The majority ofcancers seen only on US were node-negative invasive cancers; invasive lobular carcinomaand low-grade invasive ductal carcinoma were overrepresented among such cancers.

One of the major concerns about screening is the harm of extra testing and biopsies forwomen who do not have cancer.20 As has been observed with mammography21 andMRI,11,22–24 the risk of false positives decreased significantly with annual screeningultrasound in this study compared to the first screen. However, there still remained asubstantial rate of biopsies prompted only by incidence screening ultrasound, averaging5.0% (242/4814) of women, with only 7.4% (18/242) of those biopsied only due toultrasound found to have cancer.

In a separate analysis of ACRIN 6666 participants, MRI was significantly less tolerable thanmammography or ultrasound.25 Only 58% of ACRIN 6666 participants offered a screeningMRI at no out-of-pocket cost accepted the invitation.15 These barriers are in addition to highcosts of MRI equipment, contrast, and examination, as well as high rates of induced testingincluding biopsy.

Contrast-enhanced MRI has been recommended for supplemental screening of women athigh risk for breast cancer, defined as those women with a lifetime risk of 20 to 25% or

Berg et al.


NIH


NIH


NIH


greater based on family history or prior mantle radiation to the chest.26 Across nine series,the supplemental yield of MRI after mammography in high-risk women was 11 per 1000(reviewed in 27), and was 14 per 1000 among the subset who also had screeningultrasound.11–13,24 Similar results were observed in this study of women who were mostly atintermediate risk for breast cancer.

Fewer studies have evaluated MRI in women at intermediate risk, including women withPHBC, prior atypical biopsy or lobular carcinoma in situ (LCIS), intermediate family historyof breast cancer (lifetime risk of 15–20% per the ACS guidelines26), or women whose onlyrisk factor is dense breasts. Recent studies collectively suggest that supplemental MRIscreening may be reasonable for women with PHBC and also found false positives to be lessfrequent than in women with a family history of breast cancer.28–30

For high-risk women unable to undergo MRI,15 and for intermediate-risk women with densebreasts, including those with PHBC, this study supports the use of supplemental screeningwith ultrasound in addition to mammography. With either MRI or US, the risks of falsepositives, including unnecessary biopsies, were lower for supplemental screening in womenwith PHBC than in women without. The outcomes in terms of staging, node-positivedisease, and interval cancer rates achieved in this study after three years of programmaticscreening with both ultrasound and mammography were comparable to benchmarks fromstudies which included MRI.10–13,24

If screening ultrasound were to be adopted for women with dense breasts who are notcandidates for MRI, there would be obstacles to its implementation. These include theavailability of only one current procedural terminology (CPT) code, 76645, for breastultrasound, with low reimbursement (2010 Medicare reimbursement averaged a global feeof $89.85 to $91.83,31 which does not cover the costs of physicians performing andinterpreting a thorough screening examination). While supplemental cancer detection rateswith technologist-performed screening US were similar to physician-performed US in oneseries,4 there remains a shortage of qualified breast US technologists.

There are a few limitations to this study. Additional node-negative invasive cancers werefound by adding screening ultrasound to mammography in each incidence screen, andincreasing detection of such cancers correlates with mortality reduction;32 however, we didnot have a control group with no ultrasound performed in which to compare clinicaloutcomes, and mortality was not assessed. In Japan, the ongoing J-START trial of biennialmammography, with or without technologist-performed screening ultrasound, will followoutcomes to mortality reduction.33 We only performed a single screening MRI, and falsepositives would be expected to decrease in subsequent years.11,22 Not all sites in the originalACRIN 6666 protocol were able to offer MRI.

SUMMARYIn summary, the cancer detection benefit to supplemental screening ultrasound seen on thefirst screen persisted with each annual screen. Rates of biopsy for findings seen only onultrasound remained substantial on incidence screens, representing 5% of women, with only7.4% of those women found to have cancer. Risks of false positives were lower in womenwith PHBC than in women without.

As has been seen in other studies10,11,24,34, MRI significantly increased detection of earlybreast cancer beyond that seen with mammography or mammography combined withultrasound. The 56% absolute increase in cancer detection seen in the MRI substudy (16/16vs. 7/16) was greater than the 34% absolute increase in invasive cancer detection (71/89 vs.41/89) seen by adding annual ultrasound to mammography in the main ACRIN 6666 study.

Berg et al.


NIH


NIH


NIH


However, given the low (clinically-detected) interval cancer rate of only 8% in the mainACRIN 6666 protocol, and the fact that all interval cancers remained node-negative atdiagnosis, it is unclear that the added cost and reduced tolerability of screening MRI arejustified in women at intermediate risk for breast cancer in lieu of supplemental screeningwith ultrasound. Despite its higher sensitivity, addition of screening MRI rather thanultrasound to mammography in broader populations of women at intermediate risk withdense breasts may not be appropriate, particularly when the current high false positive rates,cost, and reduced tolerability of MRI are considered.

Supplementary MaterialRefer to Web version on PubMed Central for supplementary material.

AcknowledgmentsFunding/Support: The study was funded by the Avon Foundation and grants CA 80098 and CA 79778 from theNational Cancer Institute.

Role of the Sponsors: The Avon Foundation was not involved in the design or conduct of the study; collection,management, analysis, or interpretation of the data; or preparation, review, or approval of the manuscript. The trialwas conducted by the American College of Radiology Imaging Network, a member of the National CancerInstitute’s Clinical Trials Cooperative Groups Program, and was developed and carried out adhering to the standardcooperative group processes. These processes include review of and input about the trial design from the NCI’sCancer Therapy Evaluation Program (CTEP). Upon CTEP’s approval of the research protocol, the NCI was notinvolved in the design or conduct of the study; collection, management, analysis, or interpretation of the data; orpreparation, review, or approval of the manuscript.

Abbreviations used

ACRIN American College of Radiology Imaging Network.

References1. Buchberger W, Niehoff A, Obrist P, DeKoekkoek-Doll P, Dunser M. Clinically and

mammographically occult breast lesions: detection and classification with high-resolutionsonography. Semin Ultrasound CT MR. 2000; 21(4):325–336. [PubMed: 11014255]

2. Crystal P, Strano SD, Shcharynski S, Koretz MJ. Using sonography to screen women withmammographically dense breasts. AJR Am J Roentgenol. 2003 Jul; 181(1):177–182. [PubMed:12818853]

3. Gordon PB, Goldenberg SL. Malignant breast masses detected only by ultrasound. A retrospectivereview [see comments]. Cancer. 1995; 76(4):626–630. [PubMed: 8625156]

4. Kaplan SS. Clinical utility of bilateral whole-breast US in the evaluation of women with densebreast tissue. Radiology. 2001 Dec; 221(3):641–649. [PubMed: 11719658]

5. Kolb TM, Lichy J, Newhouse JH. Comparison of the performance of screening mammography,physical examination, and breast US and evaluation of factors that influence them: An analysis of27,825 patient evaluations. Radiology. 2002 Oct; 225(1):165–175. [PubMed: 12355001]

6. Leconte I, Feger C, Galant C, et al. Mammography and subsequent whole-breast sonography ofnonpalpable breast cancers: the importance of radiologic breast density. AJR Am J Roentgenol.2003 Jun; 180(6):1675–1679. [PubMed: 12760942]

7. Berg WA, Blume JD, Cormack JB, et al. Combined screening with ultrasound and mammographyvs mammography alone in women at elevated risk of breast cancer. JAMA. 2008 May 14; 299(18):2151–2163. [PubMed: 18477782]

8. Corsetti V, Ferrari A, Ghirardi M, et al. Role of ultrasonography in detecting mammographicallyoccult breast carcinoma in women with dense breasts. Radiol Med (Torino). 2006 Apr; 111(3):440–448. [PubMed: 16683089]

Berg et al.


NIH


NIH


NIH


9. Tohno E, Ueno E, Watanabe H. Ultrasound screening of breast cancer. Breast Cancer. 2009; 16(1):18–22. [PubMed: 19009372]

10. Kuhl C, Weigel S, Schrading S, et al. Prospective multicenter cohort study to refine managementrecommendations for women at elevated familial risk of breast cancer: the EVA trial. J ClinOncol. 2010 Mar 20; 28(9):1450–1457. [PubMed: 20177029]

11. Kuhl CK, Schrading S, Leutner CC, et al. Mammography, breast ultrasound, and magneticresonance imaging for surveillance of women at high familial risk for breast cancer. J Clin Oncol.2005 Nov 20; 23(33):8469–8476. [PubMed: 16293877]

12. Lehman CD, Isaacs C, Schnall MD, et al. Cancer yield of mammography, MR, and US in high-riskwomen: prospective multi-institution breast cancer screening study. Radiology. 2007 Aug; 244(2):381–388. [PubMed: 17641362]

13. Sardanelli F, Podo F, D'Agnolo G, et al. Multicenter comparative multimodality surveillance ofwomen at genetic-familial high risk for breast cancer (HIBCRIT study): interim results.Radiology. 2007 Mar; 242(3):698–715. [PubMed: 17244718]

14. D'Orsi, CJ.; Bassett, LW.; Berg, WA., et al. Breast Imaging Reporting and Data System, BI-RADS: Mammography. 4th edition. Reston: American College of Radiology; 2003.

15. Berg WA, Blume JD, Adams AM, et al. Reasons women at elevated risk of breast cancer refusebreast MR imaging screening: ACRIN 6666. Radiology. 2010 Jan; 254(1):79–87. [PubMed:20032143]

16. Leemis LM, Trivedi KS. A comparison of approximate interval estimators for the bernoulliparameter. Amer Statistician. 1996; 50:63–68.

17. Fleiss, JL.; Levin, B.; Paik, MC. Statistical Methods for Rates and Proportions. Hoboken: Wiley-Interscience; 2003.

18. Efron B. Bootstrap Methods: Another Look at the Jackknife. Ann Stat. 1979; 7(1):1–26.

19. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlatedreceiver operating characteristic curves: a nonparametric approach. Biometrics. 1988 Sep; 44(3):837–845. [PubMed: 3203132]

20. Nelson HD, Tyne K, Naik A, Bougatsos C, Chan BK, Humphrey L. Screening for breast cancer: anupdate for the U.S Preventive Services Task Force. Ann Intern Med. 2009 Nov 17; 151(10):727–737. W237–742. [PubMed: 19920273]

21. Schell MJ, Yankaskas BC, Ballard-Barbash R, et al. Evidence-based target recall rates forscreening mammography. Radiology. 2007 Jun; 243(3):681–689. [PubMed: 17517927]

22. Kriege M, Brekelmans CT, Boetes C, et al. Differences between first and subsequent rounds of theMRISC breast cancer screening program for women with a familial or genetic predisposition.Cancer. 2006 Jun 1; 106(11):2318–2326. [PubMed: 16615112]

23. Leach MO, Boggis CR, Dixon AK, et al. Screening with magnetic resonance imaging andmammography of a UK population at high familial risk of breast cancer: a prospective multicentrecohort study (MARIBS). Lancet. 2005 May 21–27; 365(9473):1769–1778. [PubMed: 15910949]

24. Warner E, Plewes DB, Hill KA, et al. Surveillance of BRCA1 and BRCA2 mutation carriers withmagnetic resonance imaging, ultrasound, mammography, and clinical breast examination. Jama.2004 Sep 15; 292(11):1317–1325. [PubMed: 15367553]

25. Schleinitz MD, De Palo DM, Blume JD, Cormack JB, Pisano ED, Berg WA. Tolerability of breastcancer screening, diagnostic, and biopsy procedures: Implications for new screening modalities.An ACRIN 6666 Substudy. Quality of Life Res. 2011 submitted.

26. Saslow D, Boetes C, Burke W, et al. American cancer society guidelines for breast screening withMRI as an adjunct to mammography. CA Cancer J Clin. 2007 Mar-Apr;57(2):75–89. [PubMed:17392385]

27. Berg WA. Tailored supplemental screening for breast cancer: what now and what next? AJR Am JRoentgenol. 2009 Feb; 192(2):390–399. [PubMed: 19155400]

28. Brennan ME, Houssami N, Lord S, et al. Magnetic resonance imaging screening of thecontralateral breast in women with newly diagnosed breast cancer: systematic review and meta-analysis of incremental cancer detection and impact on surgical management. J Clin Oncol. 2009Nov 20; 27(33):5640–5649. [PubMed: 19805685]

Berg et al.


NIH


NIH


NIH


29. Brennan S, Liberman L, Dershaw DD, Morris E. Breast MRI screening of women with a personalhistory of breast cancer. AJR Am J Roentgenol. 2010 Aug; 195(2):510–516. [PubMed: 20651211]

30. Demartini, WB.; Kalish, GM.; Peacock, S.; Eby, PR.; Gutierrez, RL.; Lehman, CD. ScreeningMRI for high risk women: should patients with a treated personal history of breast cancer bescreened?; Paper presented at: Radiologic Society of North America; November 28, 2010;Chicago, IL. 2010.

31. [accessed September 8, 2010] http://www.cms.gov/PFSlookup/01_Overview.asp#TopOfPage

32. Smith RA, Duffy SW, Gabe R, Tabar L, Yen AM, Chen TH. The randomized trials of breastcancer screening: what have we learned? Radiol Clin North Am. 2004 Sep; 42(5):793–806. v.[PubMed: 15337416]

33. Ohuchi N, Suzuki A, Sakarai Y, et al. Current Status and Problems of Breast Cancer Screening.JMAJ. 2009; 52:45–49.

34. Sardanelli F, Podo F, Santoro F, et al. Multicenter surveillance of women at high genetic breastcancer risk using mammography, ultrasonography, and contrast-enhanced magnetic resonanceimaging (the high breast cancer risk Italian 1 study): final results. Invest Radiol. 2011 Feb; 46(2):94–105. [PubMed: 21139507]

Berg et al.


NIH


NIH


NIH


http://www.cms.gov/PFSlookup/01_Overview.asp#TopOfPage

Figure 1.Flowchart of protocol. Participants with negative results on both mammography andultrasound were imputed as having negative results on integrated reading: 1844 in year one;1922 in year two; 1912 in year three. Reference standard was the most severe of biopsyresults within 365 days of mammographic screening and/or clinical follow-up at one year.Biopsies prompted by an early subsequent screening examination were attributed to thatsubsequent screen.

Berg et al.


NIH


NIH


NIH


NIH


NIH


NIH


Berg et al.

Tabl

e 1

Part

icip

ant C

hara

cter

istic

s

Ana

lysi

s Se

tSc

reen

1(N

=265

9)

Ana

lysi

s Se

tSc

reen

2(N

=249

3)

Ana

lysi

s Se

tSc

reen

3(N

=232

1)

MR

Ana

lysi

sSe

t(N

=612

)

Age

at s

can

(yrs

)M

edia

n (r

ange

)55

.0 (

25–9

1)56

.0 (

26–9

2)57

.0 (

27–9

3)57

.0 (

27–8

7)

Mea

n (S

D)

55.2

(10

.1)

56.4

(9.

9)57

.7 (

9.8)

56.8

(9.

5)

Age

gro

up a

t sca

n, n

(%

)A

ge <

40

134

(5.0

)89

(3.

6)65

(2.

8)17

(2.

8)

40 ≤

Age

≤ 4

962

7 (2

3.6)

514

(20.

6)39

2 (1

6.9)

114

(18.

6)

50 ≤

Age

≤ 6

916

78 (

63.1

)16

44 (

65.9

)15

97 (

68.8

)42

9 (7

0.1)

Age

> 6

922

0 (8

.3)

246

(9.9

)26

7 (1

1.5)

52 (

8.5)

Rac

e or

Eth

nici

ty, n

(%

)W

hite

2467

(92

.8)

2316

(92

.9)

2162

(93

.1)

576

(94.

1)

His

pani

c or

Lat

ino

265

(10.

0)23

3 (9

.3)

209

(9.0

)83

(13

.6)

Bla

ck o

r A

fric

an A

mer

ican

91 (

3.4)

85 (

3.4)

77 (

3.3)

11 (

1.8)

Nat

ive

Haw

aiia

n or

oth

er P

acif

ic I

slan

der

4 (0

.2)

3 (0

.1)

4 (0

.2)

1 (0

.2)

Asi

an90

(3.

4)82

(3.

3)71

(3.

1)22

(3.

6)

Am

eric

an I

ndia

n or

Ala

skan

Nat

ive

4 (0

.2)

4 (0

.2)

4 (0

.2)

1 (0

.2)

Unk

now

n11

(0.

4)11

(0.

4)11

(0.

5)1

(0.2

)

Men

opau

sal S

tatu

s, n

(%

)Pr

emen

opau

sala

609

(22.

9)55

4 (2

2.2)

502

(21.

6)15

5 (2

5.3)

Peri

men

opau

salb

182

(6.8

)17

0 (6

.8)

158

(6.8

)37

(6.

0)

Post

men

opau

salc

1362

(51

.2)

1294

(51

.9)

1208

(52

.0)

316

(51.

6)

Surg

ical

Men

opau

se48

4 (1

8.2)

454

(18.

2)43

2 (1

8.6)

103

(16.

8)

Unk

now

n22

(0.

8)21

(0.

8)21

(0.

9)1

(0.2

)

Pers

onal

His

tory

of

Bre

ast C

ance

r (R

egar

dles

s of

Oth

er R

isk

Fact

ors)

, n (

%)

Yes

d14

26 (

53.6

)13

31 (

53.4

)12

53 (

54.0

)27

5 (4

4.9)

Vis

ually

-Est

imat

ed B

reas

t Den

sity

at s

can,

n (

%)

<25

%47

(1.

8)47

(1.

9)34

(1.

5)7

(1.1

)

26–4

0%27

8 (1

0.5)

236

(9.5

)19

6 (8

.4)

61 (

10.0

)

41–6

0%82

4 (3

1.0)

792

(31.

8)77

4 (3

3.3)

191

(31.

2)

61–8

0%99

4 (3

7.4)

976

(39.

1)92

0 (3

9.6)

253

(41.

3)

Gre

ater

than

80%

515

(19.

4)44

2 (1

7.7)

395

(17.

0)10

0 (1

6.3)


NIH


NIH


NIH


Berg et al.

Ana

lysi

s Se

tSc

reen

1(N

=265

9)

Ana

lysi

s Se

tSc

reen

2(N

=249

3)

Ana

lysi

s Se

tSc

reen

3(N

=232

1)

MR

Ana

lysi

sSe

t(N

=612

)

Unk

now

n1

(0.0

)0

(0.0

)2

(0.1

)0

(0.0

)

Prim

ary

Ris

k Fa

ctor

,e n

(%

)M

utat

ion

in B

RC

A-1

or

BR

CA

-223

(0.

9)20

(0.

8)18

(0.

8)3

(0.5

)

His

tory

of

Prio

r C

hest

and

/or

Med

iast

inal

and

/or

Axi

llary

Irr

adia

tion

8 (0

.3)

6 (0

.2)

6 (0

.3)

2 (0

.3)

Pers

onal

His

tory

of

Bre

ast C

ance

r14

13 (

53.1

)13

21 (

53.0

)12

44 (

53.6

)27

3 (4

4.6)

Lif

etim

e R

isk,

Gai

l/Cla

us M

odel

≥ 2

5%e

504

(19.

0)46

0 (1

8.5)

425

(18.

3)13

5 (2

2.1)

Five

-Yea

r R

isk,

Gai

l Mod

el ≥

2.5

%40

6 (1

5.3)

391

(15.

7)36

6 (1

5.8)

113

(18.

5)

Five

-Yea

r R

isk,

Gai

l Mod

el ≥

1.7

% a

nd E

xtre

mel

y D

ense

Bre

asts

225

(8.5

)21

7 (8

.7)

195

(8.4

)70

(11

.4)

AD

H/A

LH

/LC

IS o

r A

typi

cal P

apill

oma

80 (

3.0)

78 (

3.1)

67 (

2.9)

16 (

2.6)

a Prem

enop

ausa

l was

def

ined

as

last

men

stru

al p

erio

d w

ithin

pri

or 3

0 da

ys.

b Peri

men

opau

sal w

as d

efin

ed a

s la

st m

enst

rual

per

iod

> 3

0 da

ys a

nd <

12

mon

ths

prio

r.

c Post

men

opau

sal w

as d

efin

ed a

s la

st m

enst

rual

per

iod

at le

ast 1

2 m

onth

s pr

ior.

d 869/

1426

(71

.1%

) of

wom

en w

ith P

HB

C h

ad lu

mpe

ctom

y an

d ra

diat

ion

ther

apy

for

the

affe

cted

bre

ast(

s) o

n st

udy.

e Part

icip

ants

with

mul

tiple

ris

k fa

ctor

s w

ere

dete

rmin

ed to

hav

e a

prim

ary

risk

fac

tor

usin

g th

e fo

llow

ing

hier

arch

y: M

utat

ion

in B

RC

A-1

or

BR

CA

-2, h

isto

ry o

f pr

ior

ches

t and

/or

med

iast

inal

and

/or

axill

ary

irra

diat

ion,

per

sona

l his

tory

of

brea

st c

ance

r, li

fetim

e ri

sk, G

ail m

odel

≥ 2

5%, f

ive-

year

ris

k, G

ail m

odel

≥ 2

.5%

, fiv

e-ye

ar r

isk,

Gai

l mod

el ≥

1.7

% a

nd e

xtre

mel

y de

nse

brea

sts,

and

fin

ally

, pri

orbi

opsy

sho

win

g at

ypic

al d

ucta

l hyp

erpl

asia

(A

DH

), a

typi

cal l

obul

ar h

yper

plas

ia (

AL

H),

lobu

lar

1car

cino

ma

in s

itu, o

r at

ypic

al p

apill

oma.

f One

par

ticip

ant's

elig

ibili

ty is

bas

ed o

n a

reca

lcul

ated

Gai

l sco

re, w

here

the

orig

inal

sco

re w

as m

issi

ng.


NIH


NIH


NIH


Berg et al.

Tabl

e 2

Sum

mar

y of

Can

cer

Det

ectio

n an

d C

hara

cter

istic

s fo

r 26

62 U

niqu

e Pa

rtic

ipan

ts S

cree

ned

Thr

ee Y

ears

with

Mam

mog

raph

y an

d Ph

ysic

ian-

Perf

orm

edU

ltras

ound

and

612

Par

ticip

ants

Scr

eene

d al

so w

ith M

RI

Seen

on

Mam

mog

raph

y on

lySe

en o

n M

amm

ogra

phy

and

US

Seen

on

US

only

bef

ore

MR

IN

ot s

een

onim

agin

gSe

en o

n M

RI

only

Tot

al

Num

ber

of p

artic

ipan

ts26

6226

6226

62N

A61

2N

A

Num

ber

of s

cree

ns74

7374

7374

73N

A61

2N

A

Num

ber

of c

ance

rs33

2632

119

111

Inva

sive

can

cers

, n (

%)

18 (

55)

23 (

88)

30 (

94)

10 (

91)

8 (8

9)89

(80

)

Med

ian

size

inva

sive

tum

or, m

m (

rang

e)11

.5 (

1–55

)16

.0 (

3–0)

10.0

(2–

40)

8.5

(2–1

3)8.

5 (1

–25)

12.0

(1–

55)

Nod

al s

tagi

ng a

vaila

blea

1515

276

770

Nod

e po

sitiv

e, n

(%

)5(

33)

7(47

)1(

4)0(

0)0(

0)13

(19)

Can

cer

Typ

e an

d gr

ade,

n (

%)

ID

C17

(52

)16

(62

)24

(75

)8

(73)

7 (7

8)72

(65

)

H

igh

grad

e7

(21)

4 (1

5)6

(19)

2 (1

8)2

(22)

b21

(19

)

I

nter

med

iate

gra

de6

(18)

8 (3

1)7

(22)

1 (9

)1

(11)

23 (

21)

L

ow g

rade

3 (9

)4

(15)

11 (

34)

3 (2

7)4

(44)

25 (

23)

G

rade

can

not b

e as

sess

ed1

(3)

00

2 (1

8)0

3 (3

)

IL

C1

(3)

5 (1

9)c

5 (1

6)1

(9)

012

(11

)

Mix

ed I

DC

and

IL

C0

2 (8

)d1

(3)d

1 (9

)1

(11)

5 (5

)

DC

IS15

(45

)3

(12)

2 (6

)1

(9)

1 (1

1)22

(20

)

H

igh

grad

e2

(6)

01

(3)

1 (9

)0

4 (4

)

I

nter

med

iate

gra

de11

(33

)3

(12)

1 (3

)0

015

(14

)

L

ow g

rade

2 (6

)0

00

1 (1

1)3

(3)

Abb

revi

atio

ns u

sed:

ID

C =

inva

sive

duc

tal c

arci

nom

a; I

LC

= in

vasi

ve lo

bula

r ca

rcin

oma;

DC

IS =

duc

tal c

arci

nom

a in

situ

; gra

de w

as c

olle

cted

onl

y fo

r ID

C a

nd D

CIS

a Axi

llary

nod

al s

tatu

s co

uld

not b

e as

sess

ed f

or 1

4 pa

rtic

ipan

ts w

ith P

HB

C w

here

nod

es h

ad p

revi

ousl

y be

en r

emov

ed, n

or f

or o

ne w

ith a

per

sona

l his

tory

of

Hod

gkin

’s d

isea

se a

nd p

rior

nod

al tr

eatm

ent.

Nod

e st

atus

was

not

det

erm

ined

for

one

par

ticip

ant o

ver

age

80 a

s it

wou

ld n

ot a

ffec

t tre

atm

ent p

lann

ing.

For

thre

e pa

rtic

ipan

ts w

ithou

t nod

al s

tagi

ng, t

he r

easo

n w

as u

nkno

wn.

b Incl

udes

one

T1m

ic tu

mor

, with

gra

de b

ased

on

the

DC

IS g

rade

c Incl

udes

one

IL

C w

ith D

CIS

for

whi

ch g

rade

of

the

ILC

is m

issi

ng


NIH


NIH


NIH


Berg et al. d In

clud

es o

ne m

ixed

ID

C-I

LC

with

ass

ocia

ted

inte

rmed

iate

nuc

lear

gra

de D

CIS


NIH


NIH


NIH


Berg et al.

Tabl

e 3

Scre

enin

g Pe

rfor

man

ce in

266

2 U

niqu

e Pa

rtic

ipan

ts S

cree

ned

Thr

ee Y

ears

with

Mam

mog

raph

y an

d Ph

ysic

ian-

Perf

orm

ed U

ltras

ound

Mam

mog

raph

y (M

) al

one

Com

bine

d M

amm

ogra

phy

and

US

(M +

US)

Dif

fere

nce

of (

M +

US)

and

Mal

one

Ult

raso

und

(US)

alo

ne

Est

(No/

Tot

al)

95%

CI

Est

(No/

Tot

al)

95%

CI

Est

95%

CI

pval

ueE

st(N

o/ T

otal

)95

% C

I

Yie

ld, p

er 1

000

(N/T

otal

)

Scre

en 1

7.5

(20/

265

9)(4

.6, 1

1.6)

12.8

(34/

265

9)(8

.9, 1

7.8)

5.3

(2.1

, 8.4

)<

0.00

19.

0(2

4/ 2

659)

(5.8

, 13.

4)

Scre

en 2

3a8.

1(3

9/ 4

814)

(5.8

, 11.

1)11

.8(5

7/ 4

814)

(9.0

, 15.

3)3.

7(2

.1, 5

.8)

<0.

001

7.1

(34/

481

4)(4

.9, 9

.9)

AU

C

Scre

en 1

0.74

(0.6

3, 0

.84)

0.94

(0.8

9, 0

.99)

0.20

(0.1

0, 0

.30)

<0.

001

0.76

(0.6

6, 0

.87)

Scre

en 2

0.75

(0.6

5, 0

.86)

0.89

(0.8

2, 0

.97)

0.14

(0.0

3, 0

.25)

0.01

20.

71(0

.58,

0.8

4)

Scre

en 3

0.72

(0.6

4, 0

.81)

0.82

(0.7

4, 0

.89)

0.10

(0.0

0, 0

.18)

0.04

10.

62(0

.52,

0.7

2)

Sens

itiv

ity,

%(N

/Tot

al)

Scre

en 1

55.6

(20/

36)

(38.

1, 7

2.1)

94.4

(34/

36)

(81.

3, 9

9.3)

38.9

(20.

2, 5

7.6)

<0.

001

66.7

(24/

36)

(49.

0, 8

1.4)

Scre

en 2

352

.0(3

9/ 7

5)(4

0.2,

63.

7)76

.0(5

7/ 7

5)(6

4.7,

85.

1)24

.0(1

4.7,

33.

3)<

0.00

145

.3(3

4/ 7

5)(3

3.8,

57.

3)

Spec

ific

ity,

%(N

/Tot

al)

Scre

en 1

89.1

(233

7/ 2

623)

(87.

8, 9

0.3)

74.3

(195

0/ 2

623)

(72.

6, 7

6.0)

−14

.8(−

16.3

, −13

.2)

<0.

001

79.8

(209

2/ 2

623)

(78.

2, 8

1.3)

Scre

en 2

391

.3(4

325/

473

9)(9

0.4,

92.

1)84

.1(3

987/

473

9)(8

3.1,

85.

2)−

7.1

(−8.

0, −

6.3)

<0.

001

89.9

(425

8/ 4

739)

(89.

0, 9

0.7)

Rec

all R

ate,

%(N

/Tot

al)

Scre

en 1

11.5

(306

/ 265

9)(1

0.3,

12.

8)26

.6(7

07/ 2

659)

(24.

9, 2

8.3)

15.1

(13.

5, 1

6.6)

<0.

001

20.9

(555

/ 265

9)(1

9.3,

22.

5)

Scre

en 2

39.

4(4

53/ 4

814)

(8.6

, 10.

3)16

.8(8

09/ 4

814)

(15.

8, 1

7.9)

7.4

(6.6

, 8.2

)<

0.00

110

.7(5

15/ 4

814)

(9.8

, 11.

6)

PP

V1,

b %

(N/T

otal

)

Scre

en 1

6.5

(20/

306

)(4

.0, 9

.9)

4.8

(34/

707

)(3

.4, 6

.7)

−1.

7(−

3.7,

0.1

)0.

073

4.3

(24/

555

)(2

.8, 6

.4)

Scre

en 2

38.

6(3

9/ 4

53)

(6.2

, 11.

6)7.

0(5

7/ 8

09)

(5.4

, 9.0

)−

1.6

(−3.

1, −

0.2)

0.03

86.

6(3

4/ 5

15)

(4.6

, 9.1

)

Shor

t-te

rmF

ollo

w-u

p R

ate,

%(N

/Tot

al)

Scre

en 1

3.2

(84/

265

9)(2

.5, 3

.9)

13.8

(368

/ 265

9)(1

2.5,

15.

2)10

.7(9

.5, 1

1.9)

<0.

001

11.1

(296

/ 265

9)(1

0.0,

12.

4)

Scre

en 2

31.

6(7

6/ 4

814)

(1.2

, 2.0

)5.

3(2

56/ 4

814)

(4.7

, 6.0

)3.

7(3

.2, 4

.3)

<0.

001

3.9

(190

/ 481

4)(3

.4, 4

.5)

Bio

psy

Rat

e, %

(N

/Tot

al)

Scre

en 1

2.4

(65/

265

9)(1

.9, 3

.1)

10.2

(272

/ 265

9)(9

.1, 1

1.4)

7.8

(6.7

, 8.8

)<

0.00

18.

8(2

33/ 2

659)

(7.7

, 9.9

)

Scre

en 2

32.

0(9

7/ 4

814)

(1.6

, 2.5

)7.

0(3

39/ 4

814)

(6.3

, 7.8

)5.

0(4

.4, 5

.7)

<0.

001

5.5

(266

/ 481

4)(4

.9, 6

.2)

PP

V3,

c %

(N/T

otal

)

Scre

en 1

29.2

(19/

65)

(18.

6, 4

1.8)

11.4

(31/

272

)(7

.9, 1

5.8)

−17

.8(−

26.7

, −9.

3)<

0.00

19.

0(2

1/ 2

33)

(5.7

, 13.

4)

Scre

en 2

338

.1(3

7/ 9

7)(2

8.5,

48.

6)16

.2(5

5/ 3

39)

(12.

5, 2

0.6)

−21

.9(−

28.7

, −14

.7)

<0.

001

11.7

(31/

266

)(8

.1, 1

6.1)

a Scre

en 2

3 re

fers

to in

cide

nce

scre

ens

in y

ears

2 a

nd 3

(i.e

. at 1

2 an

d 24

mon

ths

afte

r st

udy

entr

y re

spec

tivel

y).

b PPV

1 is

def

ined

as

the

mal

igna

ncy

rate

am

ong

wom

en w

ith a

pos

itive

scr

eeni

ng te

st (

i.e. r

ecal

led

from

scr

eeni

ng f

or f

urth

er te

stin

g or

sho

rt-i

nter

val f

ollo

w-u

p).

c PPV

3 is

def

ined

as

the

mal

igna

ncy

rate

am

ong

wom

en w

ith a

pos

itive

scr

eeni

ng te

st w

ho u

nder

wen

t bio

psy

of th

e sa

me

lesi

on.


NIH


NIH


NIH


Berg et al.

Tabl

e 4

Scre

enin

g Pe

rfor

man

ce in

612

Par

ticip

ants

Scr

eene

d w

ith M

RI

afte

r T

hree

Ann

ual S

cree

ns w

ith M

amm

ogra

phy

and

Ultr

asou

nd

(M+U

S+M

R)

- (M

+US)

(M+M

RI)

- M

M+U

SM

+US+

MR

Est

95%

CI

PM

M +

MR

IE

st95

% C

IP

MR

I

Yie

ld, p

er 1

000

(No/

Tot

al)

(95%

CI)

11.4

(7/6

12)

(4.6

, 23.

4)

26.1

(16/

612)

(15.

0, 4

2.1)

14.7

(3.5

, 25.

9)0.

004

8.2

(5/6

12)

(2.7

, 19.

0)

26.1

(16/

612)

(15.

0, 4

2.1)

18.0

(5.8

, 30.

1)<

.001

22.9

(14/

612)

(12.

6, 3

8.1)

AU

C(9

5% C

I)0.

69(0

.55,

0.8

3)0.

95(0

.91,

0.9

9)0.

26(0

.11,

0.4

2)<

.001

0.63

(0.4

7, 0

.78)

0.94

(0.9

0, 0

.98)

0.31

(0.1

6, 0

.46)

<.0

010.

87(0

.75,

0.9

8)

Sens

itiv

ity,

%(N

o/T

otal

)(9

5% C

I)

43.8

(7/1

6)(1

9.8,

70.

1)

100.

0(1

6/16

)(7

9.4,

100

.0)

56.3

(25.

7, 8

6.8)

0.00

431

.3(5

/16)

(11.

0, 5

8.7)

100.

0(1

6/16

)(7

9.4,

100

.0)

68.8

(39.

8, 9

7.7)

<.0

0187

.5(1

4/16

)(6

1.7,

98.

4)

Spec

ifit

y, %

(No/

Tot

al)

(95%

CI)

84.4

(503

/596

)(8

1.2,

87.

2)

65.4

(390

/596

)(6

1.5,

69.

3)−

19.0

(−22

.3, −

15.6

)<

.001

92.1

(549

/596

)(8

9.7,

94.

1)

70.6

(421

/596

)(6

6.8,

74.

3)−

21.5

(−24

.9, −

18.0

)<

.001

75.7

(451

/596

)(7

2.0,

79.

1)

Rec

all R

ate,

%(N

o/T

otal

)(9

5% C

I)

16.3

(100

/612

)(1

3.5,

19.

5)

36.3

(222

/612

)(3

2.5,

40.

2)19

.9(1

6.6,

23.

3)<

.001

8.5

(52/

612)

(6.4

, 11.

0)

31.2

(191

/612

)(2

7.6,

35.

0)22

.7(1

9.2,

26.

2)<

.001

26.0

(159

/612

)(2

2.5,

29.

6)

PP

V1,

%(N

o/T

otal

)(9

5% C

I)

7.0

(7/1

00)

(2.9

, 13.

9)

7.2

(16/

222)

(4.2

, 11.

4)0.

2(−

3.8,

4.0

)0.

917

9.6

(5/5

2)(3

.2, 2

1.0)

8.4

(16/

191)

(4.9

, 13.

2)−

1.2

(−8.

0, 4

.6)

0.70

18.

8(1

4/15

9)(4

.9, 1

4.3)

Shor

t-te

rmfo

llow

up

rate

, %(N

o/T

otal

)(9

5% C

I)

4.6

(28/

612)

(3.1

, 6.5

)

19.6

(120

/612

)(1

6.5,

23.

0)15

.0(1

2.0,

18.

0)<

.001

0.5

(3/6

12)

(0.1

, 1.4

)

16.3

(100

/612

)(1

3.5,

19.

5)15

.8(1

2.8,

18.

9)<

.001

15.8

(97/

612)

(13.

0, 1

9.0)

Bio

psy

rate

, %(N

o/T

otal

)(9

5% C

I)

6.2

(38/

612)

(4.4

, 8.4

)

13.2

(81/

612)

(10.

7, 1

6.2)

7.0

(4.8

, 9.2

)<

.001

1.6

(10/

612)

(0.8

, 3.0

)

9.6

(59/

612)

(7.4

, 12.

3)8.

0(5

.7, 1

0.3)

<.0

018.

5(5

2/61

2)(6

.4, 1

1.0)

PP

V3,

%(N

o/T

otal

)(9

5% C

I)

18.4

(7/3

8)(7

.7, 3

4.3)

18.5

(15/

81)

(10.

8, 2

8.7)

0.1

(−8.

8, 8

.8)

0.98

350

.0(5

/10)

(18.

7, 8

1.3)

25.4

(15/

59)

(15.

0, 3

8.4)

−24

.6(−

51.2

, 3.7

)0.

081

23.1

(12/

52)

(12.

5, 3

6.8)

a p-va

lue

that

obs

erve

d di

ffer

ence

of

com

bine

d m

amm

ogra

phy,

US,

and

MR

I vs

. mam

mog

raph

y an

d U

S oc

curr

ed b

y ch

ance

b p-va

lue

that

obs

erve

d di

ffer

ence

of

com

bine

d m

amm

ogra

phy

and

MR

I vs

. mam

mog

raph

y al

one

occu

rred

by

chan

ce

c Yie

ld is

the

canc

er d

etec

tion

rate

d PPV

1 is

def

ined

as

the

mal

igna

ncy

rate

am

ong

wom

en w

ith a

pos

itive

scr

eeni

ng te

st (

i.e. r

ecal

led

from

scr

eeni

ng f

or f

urth

er te

stin

g or

sho

rt-i

nter

val f

ollo

w-u

p).

e PPV

3 is

def

ined

as

the

mal

igna

ncy

rate

am

ong

wom

en w

ith a

pos

itive

scr

eeni

ng te

st w

ho u

nder

wen

t bio

psy

of th

e sa

me

lesi

on.


women with elevated breast cancer risk nih public access

Documents