• breast carcinoma;
  • screening;
  • magnetic resonance imaging;
  • mammography;
  • ductal lavage;
  • BRCA;
  • cancer genetics


  1. Top of page
  2. Abstract


Intensive screening is an alternative to prophylactic mastectomy in women at high risk for developing breast carcinoma. The current article reports preliminary results from a screening protocol using high-quality magnetic resonance imaging (MRI), ductal lavage (DL), clinical breast examination, and mammography to identify early malignancy and high-risk lesions in women at increased genetic risk of breast carcinoma.


Women with inherited BRCA1 or BRCA2 mutations or women with a > 10% risk of developing breast carcinoma at 10 years, as estimated by the Claus model, were eligible. Patients were accrued from September 2001 to May 2003. Enrolled patients underwent biannual clinical breast examinations and annual mammography, breast MRI, and DL.


Forty-one women underwent an initial screen. Fifteen of 41 enrolled women (36.6%) either had undergone previous bilateral oophorectomy and/or were on tamoxifen at the time of the initial screen. One patient who was a BRCA1 carrier had high-grade ductal carcinoma in situ (DCIS) that was screen detected by MRI but that was missed on mammography. High-risk lesions that were screen detected by MRI in three women included radial scars and atypical lobular hyperplasia. DL detected seven women with cellular atypia, including one woman who had a normal MRI and mammogram.


Breast MRI identified high-grade DCIS and high-risk lesions that were missed by mammography. DL detected cytologic atypia in a high-risk cohort. A larger screening trial is needed to determine which subgroups of high-risk women will benefit and whether the identification of malignant and high-risk lesions at an early stage will impact breast carcinoma incidence and mortality. Cancer 2004. © 2004 American Cancer Society.

Between 9000 and 18,000 new diagnoses of breast carcinoma per year in the United States are associated with a genetically defined predisposition.1, 2 Mutations in BRCA1 and BRCA2 account for > 60% of inherited breast carcinoma. By age 50 years, 50% of mutation carriers will develop breast carcinoma. This observation emphasizes the need for sensitive screening strategies that begin at an early age. The only known effective intervention for preventing breast carcinoma in high-risk women is prophylactic mastectomy,3, 4 although tamoxifen may benefit some women with inherited BRCA mutations.5 Intensive screening with conventional mammography and clinical examination is recommended as an alternative to prophylactic surgery and as a complement to tamoxifen, despite concerns that this strategy may not have sufficient sensitivity to reduce breast carcinoma mortality and that ionizing radiation from mammography may promote BRCA-related breast carcinogenesis.6, 7 These issues are particularly relevant to women age < 50 years, in whom it has been estimated that the sensitivity of mammography is lower compared with the sensitivity in women ages 50–64 years. Furthermore, as more women age < 30 years are identified with deleterious mutations in susceptibility genes, the concern of yearly exposure to ionizing radiation from 4-view mammography becomes greater.

Screening mammography is recommended for women age > 50 years who are at average risk based on the mortality reduction demonstrated from 8 randomized, controlled trials that evaluated the ability of mammography to reduce breast carcinoma mortality, but this is not recommended not for women age < 50 years.8 Estimates of the sensitivity of mammography from these published trials, which included women of all ages, have ranged from 39% to 89%. After the exclusion of women age > 50 years, the sensitivity of mammography ranges from 39–66%. Many published studies have confirmed that the sensitivity of mammography increases with age.9, 10 The sensitivity of mammography may be decreased further in women who carry a deleterious mutation in BRCA1 or BRCA2; pilot studies evaluating mammography in BRCA1 and BRCA2 carriers have found a high rate of false-negative results.6, 11, 12 This may be attributable both to increased breast density in young women and to tumor phenotype, including features such as pushing margins, which may contribute to a smooth appearance rather than a spiculated appearance of a mass on mammography.13 The data do not support reliance on annual mammography as the sole mode for the detection of early breast carcinoma in BRCA1 and BRCA2 mutation carriers.

Contrast-enhanced magnetic resonance imaging (MRI) has demonstrated high sensitivity as a diagnostic tool for the detection of invasive breast carcinoma but reportedly has a specificity ranging from 37% to 97%.14–17 Breast lesions on contrast-enhanced MRI are assessed for malignancy based on morphologic and pharmacokinetic patterns of enhancement. Morphologic features of malignancy include spiculated borders and rim enhancement. Pharmacokinetic features of malignancy include a rapid rate of contrast uptake and an early rate of contrast wash-out. The specificity of MRI for invasive disease can be low, because benign and malignant breast disease can have similar enhancement patterns. Low specificity can lead to high rates of additional imaging and biopsy procedures. Recent studies from multiple institutions have evaluated the role of MRI for screening in high-risk populations.6, 11, 12, 18–25 These studies demonstrate that MRI increases the sensitivity of a screening protocol in mutation carriers and succeeds at detecting earlier-stage malignancies: all screen-detected tumors were T1 lesions.6, 11, 12 To date, no study has demonstrated a survival benefit from breast MRI screening.26

Ductal lavage (DL) is a recently developed approach to identify high-risk and occult malignant lesions in breast epithelial cells in women at high risk for developing breast carcinoma. Two prospective studies with long-term follow-up have shown independently that women with cellular atypia detected in breast ductal cytology have an approximately five-fold increased relative risk of developing breast carcinoma.27, 28 The clinical significance of identifying atypia in an increased-risk cohort is unclear but may translate into a decrease in breast carcinoma incidence if appropriate interventions are developed and implemented in the future.

The objective of the current study was to test the hypothesis that breast MRI and DL can identify earlier-stage breast carcinoma and high-risk lesions among women at increased genetic risk for breast carcinoma compared with mammography and clinical breast examination alone. To test this hypothesis, we developed and completed the pilot phase of a comprehensive screening protocol combining clinical breast examination, mammography, high-quality breast MRI, and DL to determine the rate at which both high-risk and malignant lesions are detected and to optimize a protocol for further evaluation in a larger cohort.


  1. Top of page
  2. Abstract

Patient Selection

Patients for the current study were recruited from the Stanford University Cancer Genetics Clinic (CGC; Stanford, CA). This program serves the Northern California community through the identification, genetic counseling and testing, and management of patients and families with an inherited predisposition to breast carcinoma as well as other malignancies.

The CGC provides assessment and longitudinal care for women with increased genetic risk of developing breast carcinoma. Women were prescreened by a genetic counselor and were offered genetic counseling and testing for BRCA1/BRCA2 mutations based on their pedigrees and their pretest risk, as estimated by the Claus and BRCAPRO models.29–31 Eligibility criteria included a documented BRCA1/BRCA2 mutation or a > 10% risk of developing breast carcinoma at 10 years based on the Claus model. Because the Claus model does not take into account a previous history of bilateral prophylactic salpingo-oophorectomy (BSO) or tamoxifen use, risk may have been over-estimated.30 All patients with a history of breast carcinoma were at least 1 year status post–completion of adjuvant therapy. In patients with prior breast carcinoma who were treated with breast-conserving therapy, the previously affected breast was eligible for screening MRI. However, the treated breast was excluded from DL if the nipple appeared significantly distorted by surgery or radiation due to concerns for increased infection rates associated with lavage under these circumstances.

Screening Protocol

The comprehensive screening protocol consisted of a biannual clinical breast examination (CBE) with annual mammogram, breast MRI, and DL. The protocol for abnormality detected on CBE required a 3–4-month follow-up examination or biopsy, as determined by clinical features. The protocol for abnormal MRI or mammogram required a 6-month follow-up or biopsy, as determined by radiographic features. The protocol for atypical cells on DL required a 3–4-month interval follow-up DL and a 6-month follow-up MRI of the affected breast, even if the previous MRI had been normal. Enrollment began in September 2001, and reported accrual ended in May 2003. Patients were required to be age ≥ 25 years or 5 years younger than the earliest age at which a relative was diagnosed with breast carcinoma. Informed consent was obtained from all patients. Initially, all examinations had to be completed within 8 weeks of each other. The protocol was revised as of November 2002 and now requires the completion of all examinations within a 2-week period. Initially, outside mammograms were allowed. For the past year, all mammograms have been done at Stanford University Hospitals and Clinics for the purposes of quality control and standardization.

MR Protocol

The breast MRI screening protocol followed the acquisition and interpretation criteria established for diagnostic breast MRI at Stanford. From September 1997 to December 2002, 1522 diagnostic MRI scans were performed on 959 patients according to this protocol. Because this protocol was optimized as a unilateral breast examination, women underwent 2 separate breast MRI examinations, 1 for each breast, 1–3 days apart and timed according to menstrual cycle. The MRIs were obtained on a 1.5-tesla imager (Signa LX; General Electric Medical Systems, Milwaukee, WI) using a dedicated, 4-coil, phased-array breast coil (MRI Devices, Waukesha, WI). Single-breast, sagittal MRI included T1 spin-echo and fat-suppressed, T2-weighted images (fast spin-echo: TR, 5000–6000; effective TE, 80–100 msec; matrix, ≥ 256 × 192 pixels; slice thickness, ≤ 4.5 mm) as well as contrast-enhanced images. Contrast-enhanced images were obtained using an interleaved rapid dynamic and high-spatial-resolution protocol designed to capture the rate of enhancement as well as the morphologic features of any enhancing lesion during and after a single-bolus, antecubital, intravenous injection of 0.1 mmol/kg gadolinium contrast agent.32 Initial dynamic imaging consisted of whole-breast, 3-dimensional spiral images33 repeated every 10.68 seconds during bolus injection (TR, 38; TE, 12.3 msec; flip angle, 40 degrees; spiral interleaves, 20; field of view [FOV], 20 cm; matrix, 188 × 188 pixels; sagittal slices, 20; slice thickness, 4.5–6.0 mm; total dynamic scan duration, 213 seconds with injection at 40 seconds after the start of the scan). High-spatial-resolution imaging was performed with centric-encoded 3DSSMT34 (TR, 33; TE, 9 msec; flip angle, 50 degrees; FOV, 20 cm; matrix, 512 × 192 pixels; number of slices, 63; slice thickness, 1.5–2.0 mm; with water-selective, spectral-spatial excitation on resonance 1-2-1 MT suppression pulse). An additional 260 seconds of dynamic imaging were performed immediately after the high-resolution scan to capture the rate of wash-out of contrast from the breast.

The exact criteria for interpreting the breast MRI examination were tailored to each patient's history and imaging findings. In general, focally enhancing lesions that measured ≥ 5 mm prompted follow-up MRI, initially at 6 months, to ensure stability. Dominant lesions that measured ≥ 5 mm with either suspicious morphologic features (e.g., rim enhancement, spiculation, linear-branching enhancement34) or suspicious dynamic enhancement features (e.g., rapid, intense, initial enhancement followed by an abrupt transition to a stable plateau signal intensity or an abrupt transition to a decreasing wash-out signal intensity35) were biopsied. The radiologist (B.L.D.) was not blinded to the genetic status of the patient.

DL Protocol

A topical anesthetic of 1% lidocaine cream was applied to the nipple approximately 20 minutes prior to the procedure. An initial attempt was made to identify fluid-yielding ducts by nipple aspiration using a suction device. If fluid-yielding ducts were not seen by this technique, then an attempt was made to lavage any nonfluid-yielding duct that could be accessed easily using a dilator, which was coated in a 1% xylocaine gel. After a duct was identified, the dilator was replaced by a lavage catheter (Cytyc Health Corporation, Boxborough, MA) through which 3–5 mL of 1% lidocaine was injected. Approximately 15 mL of 0.9% saline were injected through the afferent port of the catheter in aliquots of 2–3 mL, with breast massage and subsequent fluid collection through the efferent port after each aliquot, until 10 mL of fluid were collected through the efferent port. Characteristics of the fluid collected, including color, consistency, and presence or absence of bubbles, were described. After lavage of each duct, a 2 cm length of knotted proline was inserted into each cannulated duct, marked at the 12 o'clock (cephalad) position adjacent to the nipple, and then photographed to document ductal orifice locations. The location of the lavaged duct was recorded on a grid and was assigned a standardized letter and number to mark its position for future, repeat lavage. A histopathologic diagnosis of normal, insufficient cellular material for diagnosis (ICMD), mild atypia, marked atypia, or malignant cells was given to each sample.


  1. Top of page
  2. Abstract

Patient Characteristics

Patient characteristics are listed in Table 1. The total number of patients screened was 41. The median patient age was 42.5 years (range, 27–72 years), with the next oldest patient age 60 years. Twenty-four patients (58.5%) had a known deleterious mutation in BRCA1 or BRCA2. The other 17 patients had a > 10% risk of developing breast carcinoma at 10 years based on the Claus model or had a significant family history indicating an autosomal-dominant pattern of inheritance of a breast carcinoma susceptibility allele. One enrolled patient had Cowden disease, which was documented by clinical criteria, and four patients had BRCA variants of unknown significance. Twelve patients (29.3%) had a previous history of breast carcinoma, and 3 patients (7.3%) had a previous history of ovarian malignancy. Eleven patients (26.8%) had undergone a prior BSO, and 6 patients (14.6%) had been receiving tamoxifen for at least 6 months before their first screen or had recently completed ≥ 2 years of tamoxifen therapy; 2 patients in this group were receiving tamoxifen and had undergone a BSO, for a total of 15 patients (36.6%) who had their risk reduced. Six BRCA1 mutation carriers and two BRCA2 mutation carriers had undergone a previous BSO, and one BRCA1 carrier had been receiving tamoxifen before the initial screen. Taken together, this cohort represents a high-risk but heterogeneous group.

Table 1. Cohort Characteristics
CharacteristicNo. of patients (%)BSOTamoxifenaHistory of breast carcinomaHistory ovarian malignancy
  • BSO: bilateral salpingo-oophorectomy.

  • a

    Two patients previously underwent BSO and also were receiving tamoxifen.

  • b

    This group included 1 patient with Cowden disease (based on clinical features) and 12 patients who had a > 10% risk of breast carcinoma at 10 years (based on the Claus model).

  • c

    This group included mutations in BRCA1 and BRCA2 that have not been shown definitively to cause malignancy.

BRCA1 carrier19 (46.3)6162
BRCA2 carriers 5 (12.2)2010
Non-BRCA mutation carriersb13 (31.7)1221
Variant of unknown significancec 4 (9.8)2330
Total41 (100.0)116123

Imaging Results

Forty-one women underwent an initial round of screening with CBE, mammography, and MRI. Abnormal MRIs were seen in 25 patients, resulting in a recommendation of either a 6-month follow-up MRI (n = 14 [34.1%]; 95% confidence interval, 19.6–48.6%) or a biopsy (n = 11 [26.8%]; 95% confidence interval, 13.3–40.4%). Of the 14 patients who received recommendations for a 6-month follow-up MRI, 1 patient also underwent biopsy because of a palpable lesion that was not seen on MRI, which was benign. Another 1 of those 14 patients also underwent a contralateral 6-month follow-up MRI for atypia on DL, despite a normal MRI of that breast: both follow-up MRIs were unchanged (Patient 4) (Table 2). Of the 25 patients who had abnormal initial MRI findings, to date, 16 have undergone 6-month follow-up MRIs, resulting in no biopsies. However, 3 of those patients required repeat 6-month follow-up MRIs due to new areas of enhancement.

Table 2. Characteristics of Patients with Atypical Initial Ductal Lavage Results
Patient no.MutationFirst DLFirst mammogramFirst MRISecond DLSecond mammogramSecond MRI
  • DL: ductal lavage; MRI: magnetic resonance image; L: left; R: right; ICMD: insufficient cellular material for diagnosis; DCIS: ductal carcinoma in situ.

  • a

    Affected relatives without mutation.

  • b

    Affected relative with variant of unknown significance.

  • c

    This patient underwent bilateral mastectomies.

  • d

    These patients underwent prior bilateral salpingo-oophorectomy. No patients with atypia on DL had received tamoxifen.

  • e

    This patient had a 6-month follow-up MRI for atypia on DL despite a normal initial MRI of the left breast.

  • f

    This patient underwent prior left mastectomy for breast carcinoma, thus, all studies were on the right breast only.

  • g

    This patient had significant left breast edema from prior radiation for breast carcinoma, precluding a DL attempt.

  • h

    Biopsy revealed scar.

  • i

    Both biopsies revealed atypical lobular hyperplasia.

1Not testedaL: 2 benign; R: 1 atypical, 1 ICMDL: normal; R: normalL + R: patchy, enhancement, 6-month follow-up recommendedL: pending; R: 2 benign (same ducts)PendingPending
2NoneL: 1 atypical; R: 1 ICMDL: focal density, negative ultrasound R: dense tissueL + R: heterogeneous T1 and T2 signals, 6-month follow-up recommendedL: benign (same duct) R: pendingL + R: dense tissueL: scattered, increased enhancement; R: 2 foci < 1 cm, apparently benign
3Not testedbL: 1 atypical; R: 1 benignL + R: dense tissueL: scattered foci, apparently benign; R: 9 mm× 3 mm focus, slow enhancement, 6-month follow-up recommendedL: 1 atypical; (same duct); R: not donecNot donecL + R: sta ble
4BRCA1deL: 1 atypical; R: 1 atypicalL + R: normalL: 3.5 mm nodule, 6-month follow-up recommended; R: normalPendingPendingL + R: stable
5BRCA1fR: 1 atypicalR: normalR: numerous foci, ill-defined; 6-month follow-up recommendedR: 2 ICMD (same duct; not accessible)R: normalR: scattered foci, 2 new areas; 6-month follow-up recommended
6BRCA1dgR: 1 atypical, 2 benignL: postradiation change; R: normalL: enhancement near operative bed; 6-month follow-up recommendedPendingPendingL: patchy enhancing areas, biopsy recommendedh
    R: normal   
7NoneL: atypical; R: none accessedL: lower-outer microcalcifications, biopsy recommendedi, R: normalL: 5 mm × 8 mm upperouter focus, biopsy recommended; R: normalPendingL: postbiopsy changes; R: normalL: multiple foci, appa- rently benign; R: stable

Three of 25 patients who had abnormal MRIs had a prior history of malignancy in that breast and had previously undergone lumpectomy; 1 patient had a history of breast carcinoma in the opposite breast that was treated by mastectomy. Of the 3 patients status post lumpectomy with abnormal ipsilateral MRIs, 2 patients had stable 6-month follow-up MRIs, and the third patient underwent an MRI-prompted biopsy that revealed a fibrous scar (Patient 6) (Table 3) Among the cohort with initially normal MRIs, 15 patients to date have undergone a second screen, which has not resulted in any biopsies or 6-month follow-up MRIs.

Table 3. Characteristics of Patients who Underwent Image-Prompted Biopsy
Patient no.MutationLesion on MRIPathology resultDuctal lavage resultFollow-up imaging
  • MRI: magnetic resonance image; R: right; L: left; DCIS: ductal carcinoma in situ; ALH: atypical lobular hyperplasia; ICMD: insufficient cellular material for diagnosis.

  • a

    These patients underwent bilateral salpingo-oophorectomy before screening. No patients who underwent biopsy had received tamoxifen.

  • b

    This patient underwent previous lumpectomy with radiation of the left breast for breast carcinoma. No lavage was performed on the left breast, due to scarring.

  • c

    This patient underwent two biopsies of the left breast, one for a lesion that was seen only on MRI and the other for a lesion that was seen on a mammogram: both lesions exhibited atypical lobular hyperplasia.

1BRCA1R: 2 foci, enhancementR: radial scarNone accessedR: MRI, stable
2BRCA1L: 8 mm lesion, rapid enhancementL: 6.9 cm high-grade DCISNot done (underwent immediate bilateral mastectomies)Not done (underwent immediate bilateral mastectomies)
3BRCA1R: focus of architectural distortion, nonspecific enhancementR: radial scar and ALHL: 3 benign; R: 3 benignPending
4aBRCA1R: focus of clumped enhancementR: fibrocystic changesL: 1 benign; R: none accessedPending
5BRCA1R: 1 cm lesion, rapid enhancementR: fibrocystic changesL: 1 benign; R: 1 ICMDR: MRI showed new area of focal enhancement; 6-month follow-up MRI recommended
6bBRCA1L: patchy enhancement extending to chest wallL: fibrous scarR: 1 atypical, 2 benignPending
7cNoneL: 5mm × 8 mm upper-outer lesion, rapid enhancement (mammogram showed lower-outer calcifications)L: atypical hyperplasia (both lesions)L: 1 atypical; R: none accessedL: MRI benign and mammogram benign (postbiopsy changes)
8aBRCA2L: focus of rapid enhancementL: stromal fibrosisL: 1 ICMD, 1 benign R: 1 benignL: MRI showed improvement in scattered foci seen previously
9BRCA2L: 1 cm lesion, abnormal enhancementL: nonspecific changesL: 1 benign; R: 1 benignL: MRI, stable
10NoneL: diffuse stippled areas, rapid enhancementL: fibrocystic changesNone accessedLost to follow-up
11NoneL: 5 mm lesion, rapid enhancement; R: 2.5 cm lesion, rapid enhancementL: 2 papillomas; R: fibrocystic changesL: none accessed; R: 1 ICMDL: MRI showed postsurgical change, nodule not seen; R: MRI showed postsurgical change, nodule not seen

Biopsy Results

Eleven patients had initial MRIs that showed abnormalities for which biopsies were recommended. All underwent MRI-guided biopsies, either by wire localization or core needle (Table 3). An 8 mm, abnormally enhancing, irregular area was screen-detected by MRI and was missed by mammography; on biopsy, this proved to be high-grade DCIS (Fig. 1). The patient went on to undergo bilateral mastectomy; the left breast, in which the lesion was detected on MRI, showed residual high-grade DCIS (total lesion size was estimated at 6.9 cm) without invasive carcinoma, and the right breast, which was removed prophylactically, showed no evidence of disease. Among the other patients who had MRI results that prompted biopsies, two biopsies from two patients were consistent with radial scars (Fig. 2), three biopsied lesions from two other patients were consistent with atypical lobular hyperplasia (ALH) (Fig. 3), and two papillomas were detected in one patient (Patient 11) (Table 3). Two patients also underwent fine-needle aspiration for palpable lesions that were not detected on imaging; in both patients, the results were benign.

thumbnail image

Figure 1. Normal mammogram and breast magnetic resonance image (MRI) showing ductal carcinoma in situ (DCIS). (A) A normal left breast mammogram. (B) A left breast MRI from the same patient demonstrating a dominant, irregular, enhancing lesion that measured 8 mm. MRI-guided biopsy of this lesion revealed high-grade DCIS with comedo necrosis that measured 6.9 cm.

Download figure to PowerPoint

thumbnail image

Figure 2. Breast magnetic resonance image (MRI) showing a radial scar. This patient had a mammogram that showed no abnormalities. The MRI demonstrates a rapidly enhancing 5 mm lesion with slight spiculation. Dynamic imaging (not shown) revealed rapid initial enhancement followed by a plateau and minimal wash-out of signal intensity. MRI-guided biopsy revealed a radial scar.

Download figure to PowerPoint

thumbnail image

Figure 3. Breast magnetic resonance image (MRI) showing atypical lobular hyperplasia The left breast MRI demonstrates a 5 mm × 8 mm enhancing lesion. This lesion demonstrated rapid wash-in on dynamic imaging (not shown). MRI-guided biopsy revealed an area of atypical lobular hyperplasia.

Download figure to PowerPoint

Three of four women with biopsy results that showed DCIS and high-risk lesions (radial scars and ALH) were BRCA1 carriers. All of these lesions, except for one instance of ALH that was detected only by mammography, were screen detected by MRI and were missed by mammography (Table 3). None of the patients with DCIS or high-risk lesions had their risk reduced with either BSO or tamoxifen. In the patients who had radial scars and ALH, risk reduction with tamoxifen was discussed as a potential therapeutic intervention but was not chosen by the patients. The positive predictive value of breast MRI at biopsy was 9.0% for DCIS and 27.3% for ALH and radial scars.

A follow-up MRI was completed on 6 of 11 women who underwent biopsy (Table 3). No further biopsies have been done on these patients. One patient had a second MRI that showed new foci of abnormal enhancement, requiring a second 6-month follow-up. The other 5 patients had follow-up MRIs that showed postbiopsy change or stability, supporting the benign diagnoses made on their initial biopsies. Five of 11 patients who underwent biopsies did not undergo follow-up MRIs for the following reasons: one woman chose to undergo bilateral prophylactic mastectomies (no tumor was detected on pathologic analysis of the specimens); another woman withdrew from the study, and 3 women have yet to undergo their scheduled 6-month follow-up MRIs.

DL Results

Of 41 women who underwent an initial screening MRI, 38 women underwent the DL procedure. Three women did not undergo any attempt at DL: one woman because high-grade DCIS had been found on MR-guided biopsy, prompting her to elect immediate bilateral mastectomy, and two women who were lost to follow-up. A catheter was inserted successfully into a duct in 30 patients; in 8 of 38 patients, an attempt was made to insert the catheter but was not successful. Seven of 30 women had atypical lavage results, and a total of 8 ducts (2 ducts from 1 patient) yielded atypical results (Fig. 4). In all seven patients, atypia was mild; no instances of frank malignancy were identified. Of the ducts that yielded atypia, five were from ducts that were nonfluid yielding on initial suction aspiration. Three of seven women with atypia were BRCA mutation carriers (Table 4). Atypia was detected in DL fluid in one woman who had a normal mammogram and MRI of that breast, although her contralateral breast had an abnormal MRI and also showed atypical DL fluid (Patient 4) (Table 2). The other six patients with atypia on DL had abnormal MRIs; six of seven patients with atypia had normal mammograms. One woman with right breast atypia had an abnormal left MRI and a fibrous scar in the left breast on biopsy, and another woman with left breast atypia had an abnormal left mammogram and MRI with ALH on biopsy of that breast (Patients 6 and 7) (Tables 2, 3). Two women with atypical results had undergone a prior BSO; none of the women with atypical results had taken tamoxifen. Twenty women had benign findings, and 3 women had ICMD on lavage (Table 4). Among patients who had their risk reduced by BSO, the rate of atypia on first lavage was 33.3% (95% confidence interval, 0–71%); among patients who did not have their risk reduced by BSO or tamoxifen, the rate of atypia on first lavage was 25% (95% confidence interval, 6–44%). Of the seven women with atypia on initial lavage, four have undergone a follow-up lavage. Two of those four women, neither with a BRCA mutation, have had benign cells from the ducts that previously showed atypia. One patient who was untested for BRCA mutation, but who had affected relatives with a variant of uncertain significance, had persistent atypia of the same duct and subsequently chose bilateral prophylactic mastectomies; the fourth patient had ICMD on repeat lavage, but the previously atypical duct could not be accessed (Patients 1–3,5) (Table 2). Six of 7 patients with atypia on initial screen have undergone follow-up MRIs: one required a biopsy, which was benign (Patient 6), and another required 6-month follow-up for a new area of enhancement (Patient 5) (Table 2), but the other 4 patients had stable findings.

thumbnail image

Figure 4. Ductal atypia. (A) High-power view of a cluster of atypical cells, notable for their heterogeneity in nuclear size and shape. (B) A cluster of atypical cells at low power.

Download figure to PowerPoint

Table 4. Ductal Lavage: Initial Screen Results
CharacteristicNo. of patients% of patients (95% CI)
  • 95% CI: 95% confidence interval; ICMD: insufficient cellular material for diagnosis.

  • a

    Some patients who underwent lavage of multiple ducts had benign findings in some ducts and ICMD in others.

  • b

    Non-fluid-yielding ducts were ducts from which no fluid appeared initially on nipple suction aspiration before attempts to cannulate the duct.

Results (n = 30)  
 ICMD310.0 (0.0–20.7)
 Benigna2066.7 (49.8–83.6)
 Atypical723.3 (8.2–38.4)
 Malignant0 0.0 (—)
All patients with atypical cytology (n = 7)
 BRCA carriers342.9 (6.2–79.6)
 Bilateral salpingo-oophorectomy228.6 (0.0–62.1)
 Tamoxifen0 0.0 (—)
All ducts that exhibited atypical cytology (n = 8)
 Non-fluid-yielding ductsb562.5 (29.0–96.0)


  1. Top of page
  2. Abstract

To the best of our knowledge, the current investigation is the first reported comprehensive (albeit pilot) screening study of CBE, mammography, breast MRI, and DL in a high-risk cohort. In 41 patients, we identified 1 high-grade, extensive DCIS in a BRCA1 mutation carrier and identified several high-risk lesions, including ALH and radial scars. All lesions except for one instance of ALH were screen-detected by MRI and were not detected by mammography. The one instance of ALH was screen-detected by mammography alone.

The current study showed a lower malignancy detection rate and a higher false-positive rate on MRI screening compared with other reports.6, 12 This may result from differences in patient populations with regard to mutation type and penetrance and with regard to risk reduction. It has been shown that BSO in carriers of BRCA1 and BRCA2 mutations reduces the risk of developing malignancy of the coelomic epithelium by up to 96%, although some series have found a slightly lower risk reduction, and it is estimated that BSO reduces the risk of breast carcinoma by approximately 50% in premenopausal BRCA1 and BRCA2 mutation carriers.36, 37 A protective effect of tamoxifen in the 50% range against contralateral breast carcinoma has been observed in both BRCA1 carriers and BRCA2 carriers, although a question remains about the ability of tamoxifen to reduce the risk of hormone receptor–negative tumors that often are associated with BRCA1.5, 38 Thirty-six point six percent of the current cohort consisted of women who had their risk reduced before their initial screen. Eleven of 41 women had undergone a previous BSO, and 6 of 41 women were taking tamoxifen: two of those women had their risk reduced by both methods. Other published screening studies only variably have included this information.6, 11, 12 It is notable that none of the patients in our series who had malignant or high-risk results on biopsy had their risk reduced, and no patient with prior risk reduction had a malignant or high-risk result on biopsy. If this trend continues, then it would confirm previous reports that BSO and tamoxifen are protective in patients who are at increased genetic risk of breast carcinoma.36–38 A larger sample size is needed to determine the degree to which different subgroups of high-risk patients will benefit from MRI screening, with particular attention to women who have undergone BSO or who are taking tamoxifen.

The clinical significance of high-risk lesions in patients at genetically increased risk of breast carcinoma remains unknown. Pathology studies in the general population report that a diagnosis of atypical hyperplasia on biopsy is associated with a 4.9–5.3-fold increase in the relative risk of breast carcinoma.39 A modification of the Gail model, which is used widely for estimating breast carcinoma risk, has included a multiplication factor to reflect the recognized contribution of a patient's prior history of atypical hyperplasia.40 Two groups have evaluated the prevalence of high-risk lesions (including ALH, atypical ductal hyperplasia, lobular carcinoma in situ, and DCIS) in prophylactic mastectomy specimens from women who were at increased genetic risk and found incidences of 46% and 57%.41, 42 Those study populations were comparable to ours with regard to age, BRCA1/BRCA2 mutation status, and prior risk reduction with tamoxifen and/or BSO. In the current study, it was found that 4 of 41 patients (9.8%) had DCIS or high-risk findings, including ALH and radial scars. Much of the data on high-risk lesions in women at increased genetic risk for breast carcinoma come from mastectomy series; therefore, it is difficult to predict the contribution of such lesions to patients' subsequent risk of developing malignant disease. It is possible that high-risk lesions are similarly predictive of future breast carcinoma in women at increased genetic risk and women in the general population and that clinical benefit will follow from earlier detection of these lesions in such patients.

Other investigators have demonstrated that mild cytological atypia on DL is identified in the affected breast of patients with known breast carcinoma and in either breast of patients who are at increased risk for breast carcinoma.43, 44 Mild atypia may suggest an unrecognized malignant lesion or may represent a high-risk lesion with the potential to become malignant. We identified 7 women with atypical cytology, 42.9% of whom were BRCA carriers, out of 30 women who underwent successful DL in our high-risk cohort. One patient with atypia had a high-risk lesion (ALH) that involved the same breast. No patient who had her risk reduced by tamoxifen had atypia. There did not appear to be much difference in the rate of atypia between patients who had their risk reduced by prior BSO and patients who had not, although the small numbers and broad confidence intervals make it difficult to estimate the true difference in rates. Five of 8 samples with atypia (62.5%) were from nonfluid-yielding ducts. The finding of atypia identified in nonfluid-yielding ducts has not been reported previously and suggests that the identification and lavage of nonfluid-yielding ducts should be investigated in clinical trials evaluating the utility of DL in risk assessment.

DL is an experimental technology, and its clinical role remains to be defined. Significant interreader variability and the use of different pathologic criteria in different series make the reliable diagnosis of cytologic atypia challenging.45 Given time limitations, two to three ducts per breast (a fraction of the known total of five to eight ducts per breast) are identified and lavaged on our current protocol. This may introduce some sampling bias: a finding of benign epithelial cells in three ducts may not be representative of the other ducts that may be atypical cytologically. In the current study, only one of seven patients with atypia had a normal MRI of the affected breast; six of seven patients with atypia had normal mammograms. In the future, atypical cells on DL may be combined with abnormal MRI results to permit earlier identification of a breast that is at higher risk for developing malignant disease. Continued studies with larger numbers of patients also are merited to determine whether MRI can reliably identify women who are not at risk for atypia on DL and, thus, potentially spare them from undergoing a procedure. Alternatively, DL may be most useful as a tool to collect breast cells for determining the patterns of gene and protein expression associated with variations in breast carcinoma risk. Longer follow-up is required to determine whether the detection of atypia in lavage fluid can predict the development of breast carcinoma accurately.

The current preliminary study suggests that MRI can identify DCIS and high-risk lesions that are missed on clinical breast examination and mammography in a high-risk population and that DL in these patients can identify atypical cells that are associated with an increased risk for breast carcinoma. The study protocol, although complex, has the potential to provide answers to many different questions. A large, multicenter trial is needed to determine the MRI interpretation criterion for screening high-risk women, the trade-off between MRI sensitivity and specificity, and the clinical utility of combining DL with MRI in a high-risk population. A collaborative effort between breast MRI centers is needed for rapid accrual of patients to answer these important questions relevant to the management of women who are at high genetic risk for developing breast carcinoma.


  1. Top of page
  2. Abstract
  • 1
    Madigan MP, Ziegler RG, Benichou J, Byrne C, Hoover RN. Proportion of breast cancer cases in the United States explained by well-established risk factors. J Natl Cancer Inst. 1995; 87: 16811685.
  • 2
    Ford D, Easton DF, Peto J. Estimates of the gene frequency of BRCA1 and its contribution to breast and ovarian cancer incidence. Am J Hum Genet. 1995; 57: 14571462.
  • 3
    Hartmann LC, Schaid DJ, Woods JE, et al. Efficacy of bilateral prophylactic mastectomy in women with a family history of breast cancer. N Engl J Med. 1999; 340: 7784.
  • 4
    Hartmann LC, Sellers TA, Schaid DJ, et al. Efficacy of bilateral prophylactic mastectomy in BRCA1 and BRCA2 gene mutation carriers. J Natl Cancer Inst. 2001; 93: 16331637.
  • 5
    King MC, Wieand S, Hale K, et al. Tamoxifen and breast cancer incidence among women with inherited mutations in BRCA1 and BRCA2: National Surgical Adjuvant Breast and Bowel Project (NSABP-P1) Breast Cancer Prevention Trial. JAMA. 2001; 286: 22512256.
  • 6
    Warner E, Plewes DB, Shumak RS, et al. Comparison of breast magnetic resonance imaging, mammography, and ultrasound for surveillance of women at high risk for hereditary breast cancer. J Clin Oncol. 2001; 19: 35243531.
  • 7
    Friedenson B. Is mammography indicated for women with defective BRCA genes? Implications of recent scientific advances for the diagnosis, treatment, and prevention of hereditary breast cancer. MedGenMed. 2000; 2: E9.
  • 8
    Smith RA, Cokkinides V, Eyre HJ. American Cancer Society guidelines for the early detection of cancer, 2003. CA Cancer J Clin. 2003; 53: 2743.
  • 9
    Houssami N, Irwig L, Simpson JM, McKessar M, Blome S, Noakes J. Sydney Breast Imaging Accuracy Study: comparative sensitivity and specificity of mammography and sonography in young women with symptoms. AJR Am J Roentgenol. 2003; 180: 935940.
  • 10
    Saarenmaa I, Salminen T, Geiger U, et al. The effect of age and density of the breast on the sensitivity of breast cancer diagnostic by mammography and ultasonography. Breast Cancer Res Treat. 2001; 67: 117123.
  • 11
    Tilanus-Linthorst MM, Obdeijn IM, Bartels KC, de Koning HJ, Oudkerk M. First experiences in screening women at high risk for breast cancer with MR imaging. Breast Cancer Res Treat. 2000; 63: 5360.
  • 12
    Kuhl CK, Schmutzler RK, Leutner CC, et al. Breast MR imaging screening in 192 women proved or suspected to be carriers of a breast cancer susceptibility gene: preliminary results. Radiology. 2000; 215: 267279.
  • 13
    Tilanus-Linthorst M, Verhoog L, Obdeijn IM, et al. A BRCA1/2 mutation, high breast density and prominent pushing margins of a tumor independently contribute to a frequent false-negative mammography. Int J Cancer. 2002; 102: 9195.
  • 14
    Harms SE, Flamig DP. MR imaging of the breast: technical approach and clinical experience. Radiographics. 1993; 13: 905912.
  • 15
    Cross M, Harms SE, Cheek JH, Peters GN, Jones RC. New horizons in the diagnosis and treatment of breast cancer using magnetic resonance imaging. Am J Surg. 1993; 166: 749753; discussion, 753–755.
  • 16
    Kaiser WA. [Magnetic resonance tomography of the breast. The results of 253 examinations]. Dtsch Med Wochenschr. 1989; 114: 13511357.
  • 17
    Orel SG. High-resolution MR imaging of the breast. Semin Ultrasound CT MR. 1996; 17: 476493.
  • 18
    Robson ME, Morris E, Kauff N, et al. Breast cancer screening utilizing magnetic resonance imaging (MRI) in carriers of BRCA mutations [abstract]. Proc Am Soc Clin Oncol. 2003; 22: 362a.
  • 19
    Kuhl CK, Schrading S, Leutner CC, et al. Surveillance of “high risk” women with proven or suspected familial (hereditary) breast cancer: first mid-term results of a multi-modality clinical screening trial [abstract]. Proc Am Soc Clin Oncol. 2003; 22: 4a.
  • 20
    Kriege M, Brekelmans CT, Boetes C, et al. MRI screening for breast cancer in women with high familial and genetic risk: first results of the Dutch MRI screening study (MRISC) [abstract]. Proc Am Soc Clin Oncol. 2003; 22: 5a.
  • 21
    Trecate G, Vergnaghi D, Bergonzi S, et al. BMRI in early detection of breast cancer in patients with increased genetic risk: our preliminary results. J Exp Clin Cancer Res. 2002; 21(3 Suppl ): 125130.
  • 22
    Trecate G, Vergnaghi D, Bergonzi S, et al. Breast MRI screening in patients with increased familial and/or genetic risk for breast cancer: a preliminary experience. Tumori. 2003; 89: 125131.
  • 23
    Podo F, Sardanelli F, Canese R, et al. The Italian multi-centre project on evaluation of MRI and other imaging modalities in early detection of breast cancer in subjects at high genetic risk. J Exp Clin Cancer Res. 2002; 21(3 Suppl ): 115124.
  • 24
    Leach MO, Eeles RA, Turnbull LW, et al. The UK national study of magnetic resonance imaging as a method of screening for breast cancer (MARIBS). J Exp Clin Cancer Res. 2002; 21(3 Suppl ): 107114.
  • 25
    Stoutjesdijk MJ, Boetes C, Jager GJ, et al. Magnetic resonance imaging and mammography in women with a hereditary risk of breast cancer. J Natl Cancer Inst. 2001; 93: 10951102.
  • 26
    Plevritis SK, Ikeda DM. Ethical issues in contrast-enhanced magnetic resonance imaging screening for breast cancer. Top Magn Reson Imaging. 2002; 13: 7984.
  • 27
    Wrensch M, Petrakis NL, King EB, Lee MM, Milke R. Breast cancer risk associated with abnormal cytology in nipple aspirates of breast fluid and prior history of breast biopsy. Am J Epidemiol. 1993; 137: 829833.
  • 28
    Fabian CJ, Kimler BF, Zalles CH. Short-term breast cancer prediction by random periareolar fine-needle aspiration cytology and the Gail risk model. J Natl Cancer Inst. 2000; 92: 12171227.
  • 29
    Berry DA, Parmigiani G, Sanchez J, Schildkraut J, Winer E. Probability of carrying a mutation of breast-ovarian cancer gene BRCA1 based on family history. J Natl Cancer Inst. 1997; 89: 227238.
  • 30
    Claus EB, Risch N, Thompson WB. The calculation of breast cancer risk for women with a first degree family history of ovarian cancer. Breast Cancer Res Treat. 1993; 28: 115120.
  • 31
    Parmigiani G, Berry DA, Aguilar O. Determining carrier probabilities for breast cancer-susceptibility genes BRCA1 and BRCA2. Am J Hum Genet. 1998; 62: 145158.
  • 32
    Agoston AT, Daniel B, Herfkens RJ, et al. Intensity-modulated parametric mapping for simultaneous display of rapid dynamic and high-spatial-resolution breast MR imaging data. Radiographics. 2001; 21: 217226.
  • 33
    Yen YF, Han K, Daniel BL, et al. Dynamic breast MRI with spiral trajectories: 3D versus 2D. J Magn Reson Imaging. 2000; 11: 351359.
  • 34
    Leong CS, Daniel BL, Herfkens RJ, et al. Characterization of breast lesion morphology with 3DSSMT: an adjunct to dynamic breast MRI. J Magn Reson Imaging. 2000; 11: 8796.
  • 35
    Daniel BL, Yen YF, Glover GH, et al. Breast disease: dynamic spiral MR imaging. Radiology. 1998; 209: 499509.
  • 36
    Rebbeck TR, Lynch HT, Neuhausen SL, et al. Prophylactic oophorectomy in carriers of BRCA1 or BRCA2 mutations. N Engl J Med. 2002; 346: 16161622.
  • 37
    Kauff ND, Satagopan JM, Robson ME, et al. Risk-reducing salpingo-oophorectomy in women with a BRCA1 or BRCA2 mutation. N Engl J Med. 2002; 346: 16091615.
  • 38
    Narod SA, Brunet JS, Ghadirian P, et al. Tamoxifen and risk of contralateral breast cancer in BRCA1 and BRCA2 mutation carriers: a case-control study. Hereditary Breast Cancer Clinical Study Group. Lancet. 2000; 356: 18761881.
  • 39
    Page DL, Dupont WD, Rogers LW, Rados MS. Atypical hyperplastic lesions of the female breast. A long-term follow-up study. Cancer. 1985; 55: 26982708.
  • 40
    Gail MH, Brinton LA, Byar DP, et al. Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. J Natl Cancer Inst. 1989; 81: 18791886.
  • 41
    Kauff ND, Brogi E, Scheuer L, et al. Epithelial lesions in prophylactic mastectomy specimens from women with BRCA mutations. Cancer. 2003; 97: 16011608.
  • 42
    Hoogerbrugge N, Bult P, de Widt-Levert LM,et al. High prevalence of premalignant lesions in prophylactically removed breasts from women at hereditary risk for breast cancer. J Clin Oncol. 2003; 21: 4145.
  • 43
    Khan SA, Rodriguez N, Baird C, et al. Ductal lavage findings in women with known breast cancer undergoing mastectomy [abstract]. Presented at the 25th Annual San Antonio Breast Cancer Symposium, San Antonio, TX, December 10–13, 2001.
  • 44
    Dooley WC, Ljung BM, Veronesi U, et al. Ductal lavage for detection of cellular atypia in women at high risk for breast cancer. J Natl Cancer Inst. 2001; 93: 16241632.
  • 45
    Sidawy MK, Stoler MH, Frable WJ, et al. Interobserver variability in the classification of proliferative breast lesions by fine-needle aspiration: results of the Papanicolaou Society of Cytopathology Study. Diagn Cytopathol. 1998; 18: 150165.