Ductal Lavage For Breast Cancer Risk Assessment

Surg Oncol Clin N Am 14 (2005) 45–68

Ductal lavage for breast cancer risk assessment Aeisha Rivers, MDa, Lisa A. Newman, MD, MPH, FACSb,* a

Department of Surgery, St. Joseph’s Hospital and Medical Center, Ann Arbor, MI, USA b Breast Care Center, University of Michigan Comprehensive Cancer Center, 1500 East Medical Center Drive, 3308 CGC, Ann Arbor, MI 48109, USA

Ductal lavage is a technology that is available as a means of refining breast cancer risk assessment in selected women who are candidates for breast cancer risk prevention strategies. The ductal lavage procedure provides a noninvasive opportunity to identify abnormal proliferative activity—in the form of cellular atypia—within the ductal system of a woman who has clinical evidence of increased breast cancer risk. This article reviews: (1) the evolution and rationale for breast ductal fluid analysis as a risk assessment strategy; (2) contemporary applications of ductal lavage as a risk assessment adjunct, including result-appropriate follow-up strategies; and (3) evidence that supports the potential for using ductal lavage in translational research endeavors. Why study breast ductal fluid in women who face increased breast cancer risk? Approximately 212,000 women are diagnosed with breast cancer in the United States annually [1]. Following breast cancer diagnosis, most women face treatment decisions that involve some degree of disfiguring surgery, chemotherapy, or radiation therapy. Although these treatments will be effective in controlling disease in most cases, risks of relapse and breast cancer mortality persist over the lifetime. Despite the earlier stage distribution that has resulted from screening mammography, more than 40,000 American women die of breast cancer each year. The magnitude of this breast cancer burden has been the driving force behind investigations of * Corresponding author. E-mail address: [email protected] (L.A. Newman). 1055-3207/05/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.soc.2004.07.004

46

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

risk reduction strategies. Availability of these strategies and characterization of their potential for adverse events has strengthened the need for accurate stratification of breast cancer risk. Although statistical models for estimation of this risk are valuable, there continues to be a need for the development of maneuvers that will ‘‘fingerprint’’ proliferative activity in the breast, either through tissue or ductal fluid analyses. Current options for breast cancer prevention include prophylactic mastectomy, prophylactic oophorectomy, and endocrine manipulation with selective estrogen receptor modulators, such as tamoxifen. The degree of risk reduction is estimated at approximately 90% and 50% for prophylactic mastectomy [2] and oophorectomy [3], respectively. Chemoprevention with tamoxifen can decrease breast cancer risk by nearly 50% [4]. Comparison of the relative risk reduction benefits that are conferred by raloxifene versus tamoxifen is underway through the National Surgical Adjuvant Breast Project P-2 phase 3 prospective randomized trial, the Study of Tamoxifen and Raloxifene. Recent findings from the Arimadex versus Tamoxifen versus the Combination trial revealed that aromatase inhibitors that are used as adjuvant therapy also result in diminution of risk for new primary breast cancer; this may lead to their evaluation in the chemoprevention setting for postmenopausal women [5]. None of these risk reduction strategies is completely effective and all are associated with well-recognized risks for potentially life-threatening or disabling adverse sequelae. Selective estrogen receptor modulator (SERM) therapy can increase the risk of uterine cancer and thromboembolic phenomena; aromatase inhibitors can increase the risk of osteoporosis; prophylactic oophorectomy places women at risk for premature menopause; and prophylactic mastectomy is disfiguring, even with the best of reconstruction techniques. Clearly, these interventions should be reserved for appropriatelyselected women who face the greatest risk of breast cancer development. Although many clinicopathologic risk factors for breast cancer have been identified (Table 1), there are few options for estimating an individual woman’s absolute risk for developing breast cancer. Currently-available individualized risk assessment tools include the Claus Model [6] and the Gail Model. The Claus Model was developed by subset analysis of several thousand participants of the Contraceptives and Steroid Hormone Study, using a breast cancer case-control statistical design. This model is most appropriate for women who have a significant risk for genetic breast cancer susceptibility because it uses history of breast or ovarian cancer in the extended family, as well as age at diagnosis, to calculate the cumulative probability for breast cancer development in the individual. The Gail Model is a mathematical tool that is based on analysis of a casecontrol subset of women who participated in the American Cancer Society’s screening mammography program—the Breast Cancer Detection and Demonstration project [14]. This model uses four breast cancer risk factors (age at menarche, parity, first degree family history of breast cancer, and

47

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68 Table 1 Risk factors for breast cancer Risk factor

Relative risk

Early menarche (\11 years) [7] Late age at first live birth ([35 years) [8] Late menopause [8] Prolonged HRT ([4 years) [9] Postmenopausal obesity [10] Increased mammographic density (25–50% of breast) [11,12] Family history [13] One first-degree relative Two first-degree relatives

1.1–1.9 1.16 1.44 1.26 1.45 2.4 1.80 2.93

Abbreviation: HRT, hormone replacement therapy.

biopsy history) to generate an overall relative risk for the individual. This factor is multiplied by the subject’s age-related baseline risk for developing breast cancer to yield an individualized estimate of absolute likelihood for being diagnosed with the disease. The Gail Model has been modified to account for ethnicity (white American versus African American) in the baseline risk; to account for increased risk associated with atypia in a past biopsy; and to calculate the risk of invasive disease only. A 5-year risk of at least 1.7% (the risk of an average 60-year-old white American woman) is considered high-risk for the purpose of eligibility to participate in chemoprevention trials and for identifying women who might benefit from risk reduction counseling. Studies that compared the Gail and Claus models showed comparable risk estimates, although the Gail model estimates tend to be slightly lower [15]. In general, the Gail model is used more widely and its accuracy was validated in several populations of white American women [16–19]. The Gail model is limited by the fact that it does not account for the paternal or extended family history of breast cancer and little is known about its accuracy in nonwhite American women. Furthermore, the statistical discriminatory accuracy of the model is modest, at best’ this indicates that although the model will identify groups of high-risk women reliably, its performance is weaker for individualized risk assessment. Rockhill et al [19] and Newman et al [20] studied the model among participants of the Nurses Health Study and the Women’s Contraceptives and Reproductive Experiences Study, respectively, and reported concordance statistics of 0.54 to 0.58. This measure suggests that given any breast cancer case and control pair, the Gail model has only slightly better than a 50% chance of yielding a higher estimate for the diseased patient. Other models that are used for breast cancer risk assessment are tailored more for the prediction of whether a woman is at risk for breast cancer that is associated with an inherited mutation in one of the breast cancer susceptibility genes [21]. Typically, these models rely on more detailed family history information, age at disease onset, and ethnicity (because of

N

Selection criteria

480

Wrensch et al, 2001 [30]

3633 (Group 1)

University of Kansas patients who had high risk by way of FH/PH breast cancer or history of atypical hyperplasia/ DCIS Group 1: UCSF BCDDP participants Group 2: UCSF volunteers General screening population Vanderbilt University patients who had biopsy-proven benign breast disease Haagensen patient population from Columbia-Presbyterian Medical Center who had biopsy-proven benign breast disease

3271 (Group 2) Wrensch et al, 1992 [27] Dupont et al, 1999 [31]

2343

Bodian et al, 1993 [32,33]

1799

9494

Method of detection

Breast cancer relative risk, (95% confidence intervals)

Median follow-up

% Atypia

45 months

12%a

Periareolar FNAs for cytology

5.02 (2.01–12.56)

21 years (group 1)

2.4%

NAF

2.4 (1.6–3.7)

9 years (group 2)

0.7%

NAF

2.8 (1.5–5.5)

12.7 years

2%

NAF

4.9 (1.7–13.9)

20 years

3.0%

Surgical biopsy specimens

3.58 (2.6–5.0)

20.6 years

19%

Surgical biopsy specimens

3.0 (1.5–6.0) (moderate/ severe atypia) 2.3 (1.6–3.4) (mild atypia)

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

Investigators Fabian et al, 2000 [29]

48

Table 2 Risk of breast cancer following diagnosis of atypia

12.9 years

3.1%

Surgical biopsy specimens

2.85 (0.34–10.28)

8.3 years

7.8%

Surgical biopsy specimens

3.0 (2.1–4.1)

17 years

3.6%

Surgical biopsy specimens

121 breast cancer cases 488 controls

Nurses Health Study participants who had cancer or biopsy-proven benign breast disease

9 years

22.3% (cases) 9.6% (controls)

Surgical biopsy specimens

McDivitt et al, 1992 [36]

433 breast cancer cases 261 controls

N/A

15.9% (cases) 10.0% (controls)

Surgical biopsy specimens

Palli et al, 1991 [37]

62 breast cancer cases 315 controls

Cancer and Steroid Hormone Study participants who had cancer or biopsy-proven benign breast disease Women from Florence, Italy breast cancer screening program

All: 5.3 (3.1–8.8) With FH of breast CA: 8.9 (4.8–17) Without FH of breast CA: 3.5 (2.3–5.5) All: 3.7 (2.1–6.8) With FH of breast CA: 7.3 (1.1–50.1) Without FH of breast CA: 3.7 (1.9–7.0) Odds ratio 2.6 (1.6–4.1)

N/A

Surgical biopsy specimens

Odds ratio 13.0 (4.1–41.7)

Krieger and Hiatt, 1992 [38]

2731

San Francisco Bay, California women who had biopsy-proven benign breast disease

16 years

17.7% (cases) 2.2% (controls) 12%

Surgical biopsy specimens

Rate ratio 7.2 (BlackChabon Score 5/severe atypia)

1053

Carter et al, 1988 [34]

16692

Dupont and Page, 1985 [26]

3303

London et al, 1992 [35]

49

(continued on next page)

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

Private practice patient population who had biopsy-proven benign breast disease BCDDP participants who had history of biopsy-proven benign breast disease History of biopsy-proven benign breast disease

Hutchinson et al, 1980 [25]

50

Table 2 (continued)

N

Selection criteria

Dupont et al, 1993 [39]

95 breast cancer cases 227 controls

Byrne et al, 2000 [40]

133 breast cancer cases 610 controls

Carpenter and Love, 2002 [41]

Initial cohort: 414 patients Respondents re: outcome 56

Breast Cancer Detection and Demonstration Project mammography screening program participants Nurses Health Study participants who had cancer or biopsy-proven benign breast disease Patients undergoing ductography

% Atypia

Method of detection

Breast cancer relative risk, (95% confidence intervals)

N/A

14.7% (cases) 4.4% (controls)

Surgical biopsy specimens

Odds ratio 4.3 (1.7–11)

N/A

25.6% (cases) 11.8% (controls) NR

Surgical biopsy specimens

Odds ratio 3.6 (2.0–6.4)

Cytology on ductogramassociated ductal lavage

2.32 (1.01–4.41)

19.7 years

Abbreviations: BCDDP, Breast Cancer Demonstration and Detection Project; CA, cancer; DCIS, ductal carcinoma in situ; FH, family history; FNA, fine needle aspirate; NAF, nipple aspirate fluid; PH, personal history; UCSF, University of California-San Francisco. a Atypia found in 12% of FNAs from single aspirates; sequential FNAs over 6 and 12 months resulted in 21% atypia prevalence for pooled samples.

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

Investigators

Median follow-up

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

51

the influence of founder effects in genetically-transmitted risk). Because only 5% to 10% of breast cancer is related to hereditary disease, these types of models are not likely to be helpful in evaluating risk among large, unselected patient populations. Alternative measures of individualized risk assessment are needed as we expand chemoprevention programs and as novel risk reduction agents are developed, and require evaluation in future clinical trials. A pure, individualized predictor would be some feature that can be measured reliably and is associated consistently with increased risk. Potential candidates include mammographic density, serologic markers, or tissue characteristics. Data on magnitude of risk that is conferred by most surrogate markers have been inconsistent and interlaboratory variability in their measurement has hindered clinical application. Reproducible and individualized, measurable features of breast cancer risk are lacking; this deficiency has motivated interest in the detection of histopathologic risk factors, such as lobular carcinoma in situ, radial scar, papillomatosis, and atypical hyperplasia. The first three lesions are uncommon and only can be detected by way of open biopsy that yields a wedge of tissue for microscopic evaluation. This leaves atypical hyperplasia as the most promising feature, because atypia can be identified cytologically and on histopathologic tissue analysis. Breast carcinogenesis often is viewed as a continuum of morphologic changes at the microscopic tissue level; atypical hyperplasia appears early in the transformation process. This model for breast tumorigenesis features the evolution of breast ductal cells from normal to hyperplastic, followed by the development of atypical hyperplasia. Accumulation of genetic abnormalities as ductal cells proceed through the cell cycle leads to the development of carcinoma in situ, and ultimately, invasive cancer [22]. Autopsy findings suggest that the prevalence of atypical hyperplasia can be estimated between 12.5% and 26%, depending on the sampling technique [23,24]. Hutchinson et al [25] and Dupont and Page [26] provided some of the initial data that documented the association between atypical hyperplasia and breast cancer risk. These retrospective analyses of outcome in several thousand women who had benign breast biopsies demonstrated relative risks of 2.85 and 5.3, respectively, in cases that were associated with atypical hyperplasia. Similarly, Wrensch et al [27] reported that women who had atypia that was detected cytologically in nipple aspirate fluid had a relative risk of 4.9 for breast cancer. As shown in Table 2, several other studies have confirmed this correlation, with relative risk estimates that average between 3 and 5, regardless of whether the atypia is detected in a nipple aspirate, needle biopsy, or in an open surgical biopsy specimen. The risk that is associated with atypia may be magnified in the presence of a family history that is positive for breast cancer. One of the unique aspects of atypia as a risk indicator is that the subsequent breast cancer risk seems to be expressed predominantly in the 5 years following detection [28]; therefore, atypia may provide some temporal measure of breast cancer risk.

52

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

The efficacy of chemoprevention is an essential element in the discussion of strategies (eg, ductal lavage) that are designed to detect breast atypia. The Surgical Adjuvant Breast and Bowel Project’s (NSABP) initial chemoprevention trial of tamoxifen versus placebo in more than 13,000 high-risk women included approximately 600 women in each arm of this study who had a history of atypical hyperplasia [4]. Participants who had atypia who were randomized to tamoxifen had a highly significant, 86% lower breast cancer incidence compared with the placebo subset; this suggested that atypia is a pattern of abnormal proliferative activity that is suppressed effectively by selective estrogen receptor modulators. Thus, atypia seems to be a valuable and reliable marker of future breast cancer risk. The strength of this association is unaffected by method of tissue acquisition and it seems to be a marker of risk that is particularly sensitive to the antiproliferative effects of chemoprevention with tamoxifen. Therefore, it is reasonable to seek a reliable and low-morbidity procedure that can identify women who harbor atypia. The patient population that is most appropriate for such an intervention are high-risk women who are ambivalent about committing to a risk reduction strategy. In this setting, the detection of atypia may facilitate this decision, although the failure to detect atypia should not be misinterpreted as indicating any decrease in the level of pre-existing risk. Conventional tissue procurement maneuvers that have been incorporated into risk assessment algorithms include FNA, core needle biopsy, and open surgical biopsy. Random FNA was studied by several investigators in an attempt to determine its feasibility as a screening tool among high-risk women. The first few reports were provided by collaborators from the University of Utah in the early 1990s [42–44]. They evaluated more than 100 women with at least two first-degree relatives who were affected by breast cancer and more than 30 control patients who had no cancer diagnosis. Each subject underwent physical examination, screening mammography, and four-quadrant fine needle breast aspirates with a 1-inch, 22-gauge needle approximately 1 cm from the areola. Evidence of proliferative breast disease was seen in 35% of the participants who had a positive family history and otherwise no physical or radiologic evidence of cancer. Only 13% of the controls revealed cytologic features of proliferative breast disease (P = .02). Using a similar technique, Fabian et al [29] performed sequential FNA biopsies on 480 high-risk women. After a median follow-up of nearly 4 years, their data revealed a strong relationship between the presence of atypia in conjunction with an elevated 10-year Gail model risk estimate and an increased short-term risk of developing breast cancer. Although the information that was obtained from studies of FNA has proved to be useful, this sampling technique is not without limitations. The procedure usually is quick and inexpensive to perform, but random sampling often yields nonreproducible findings. FNA yields specimens that are sufficient for cellular and molecular marker evaluation in an estimated

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

53

85% of attempts [45]; however, the probability of obtaining adequate cellularity decreases with increasing age. Core needle biopsy has been used in the diagnosis of breast cancer for several years. Random core needle biopsy is not used often as a method of sampling breast epithelium and reports of its efficacy are scarce. One promising study by Stoler et al [46] described an alternative method of obtaining breast epithelial samples as excess tissue for possible molecular and cytogenetic studies in patients who undergo core needle biopsies. They obtained specimens from more than 100 women who underwent core needle biopsy for lesions that were detected by mammogram. Washings of these core biopsy specimens yielded an abundance of epithelial cells. Therefore, this method may become a useful tissue procurement adjunct; however, it does not fulfill the need for a noninvasive maneuver that can be performed for pure risk assessment because the candidate patients are undergoing a diagnostic core needle biopsy as the result of an abnormal breast lesion. Similarly, a random open surgical biopsy would not be feasible or reproducible for the sole purpose of risk assessment and is reserved for histopathologic differentiation between benign and malignant breast lesions. Studies of nipple aspirates for breast cancer screening and risk assessment were initiated by investigators several decades ago; however the technique was never popularized because of the low cellular content and because chemoprevention was not available in that era which limited the clinical relevance of any high-risk findings. The earliest attempts at nipple fluid aspiration were designed with the intent of detecting cancer. In the 1950s, Papanicolaou and colleagues [47,48] first reported the use of breast massage and a hand-held pump to obtain samples of nipple fluid in women who did not have signs or symptoms of breast disease. Sartorius et al [49] applied a modified version of this technique in 1700 asymptomatic women but obtained adequate fluid for cytologic evaluation in only half of the cases. In summary, procedures, such as percutaneous needle biopsy or open surgical biopsy, that are performed in a random fashion to sample breast tissue are invasive maneuvers that would be unacceptable for pure risk assessment. Direct nipple aspirates for cytologic analysis are noninvasive, but the cellular content of these specimens frequently is low and yields a nondiagnostic specimen. The goal of ductal lavage is to provide a minimallyinvasive means of extracting ductal fluid that is cytologically-enriched [50]. The ductal lavage procedure is performed in the outpatient setting with a topical anesthetic. It involves the use of a Food and Drug Administration– approved double-lumen catheter for cannulation of any fluid-yielding nipple orifice that can be identified after application of a transparent suction cup– device to the breast. Techniques that can improve the yield of fluidproducing ducts include vigorous breast massage by the patient before the procedure; nipple dekeratinization with a mild abrasive; application of warm towels; and topical nitroglycerin to the nipple [51]. Cannulation of the duct is followed by lavage with approximately 10 to 20 mL of saline and

54

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

aspiration for cytology evaluation. The process is repeated for all additional fluid-yielding ducts, using a separate catheter for each. Technical limitations of the procedure include nipple sphincter spasm that prevents duct cannulation and duct perforation secondary to increased ductal pressures during lavage. When cannulation, lavage and specimen retrieval are successful, the position of the cannulated duct should be recorded. This can be accomplished by notation on a diagrammatic grid or by photograph, with a segment of suture material inserted into the lavaged ducts for documentation and marking. These records are particularly valuable in cases where repeat lavage is considered. A multicenter study that was reported by Dooley et al [52] confirmed the superiority of ductal lavage over direct aspirates in yielding cytologicallyevaluable fluid among high-risk women. This landmark 2001 report compared the efficacy of the two techniques in a series of more than 500 women who were identified as being at high-risk for breast cancer on the basis of family history, personal history of breast cancer, or a 5-year Gail model breast cancer risk estimate of at least 1.7%. Eighty-four percent of study participants had fluid-yielding ducts that were amenable to lavage; 82% of these fluid-yielding ducts were cannulated successfully. Among the population who tolerated cannulation for lavage, 78% had samples with adequate cellular material for diagnosis. Substantially more cells were collected with ductal lavage compared with nipple fluid aspiration (13,500 cells per duct versus 120 epithelial cells). Ductal lavage was 3.5 times more successful at producing cytologically-evaluable fluid compared with paired nipple aspirates (72% versus 21%, respectively; P \ .001) in this study. Among the women who were evaluated with ductal lavage, 92 (24%) subjects showed evidence of cellular abnormalities that were mildly (17%) or markedly (6%) atypical or malignant (\1%). In contrast, abnormal cells were detected in only 10% of the women who underwent nipple fluid aspiration; mild atypia was identified in 6%, marked atypia was identified in 3%, and malignancy was identified in less than 1% of the women who were evaluated. The collaborators concluded that ductal lavage is the more sensitive technique for detecting atypia. Generally, the ductal lavage procedure is well-tolerated by patients. In the multi-center study by Dooley et al [52], the median discomfort level that was reported by study participants was 24 on a visual analog scale of 1 to 100, comparable in magnitude to the discomfort level described for mammography. There also was minimal morbidity, with no major complications of the procedure and only two cases of suspected infections that were treated uneventfully with oral antibiotics. A frequently-cited criticism of the procedure is the fact that there are no data to confirm the magnitude of future breast cancer risk that is conferred by the detection of atypia from a ductal lavage specimen. The strength of this association is similar for all conventional diagnostic procedures (eg, tissue biopsies versus cytologic assays for nipple aspirates); therefore, it

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

55

seems reasonable to extrapolate that this consistency would extend to atypia detected by ductal lavage (see Table 2). Nonetheless, documentation of this risk based on the long-term follow-up of patients who underwent lavage remains necessary. Interesting data on the outcome of patients who underwent a variation of the ductal lavage procedure were provided in a study that was presented by Carpenter and Love [41]. This study involved an attempt to compile follow-up information on patients from the Sartorius breast practice [49] who underwent ductography with concomitant lavage through the ductogram catheter more than 30 years ago. The power of this study is limited by the poor follow-up success. The original cohort consisted of 414 patients; however, there were only 56 respondents out of 191 patients (29%) who were believed to be alive at the nearly 20-year median follow-up. Nonetheless, the investigators confirmed that atypia that was detected by this modified lavage procedure conferred a statistically significant greater than twofold relative risk for breast cancer. As a risk assessment adjunct, ductal lavage also would be expected to identify candidates for chemoprevention trial eligibility. The presence of atypical hyperplasia provides more concrete and persuasive evidence of breast cancer risk to candidates for chemoprevention compared with conventional measures of risk, such as the Gail model; this suggests that ductal lavage can be a powerful tool in this regard. Port et al [53] reported the experience of 43 high-risk women who were seen at the Memorial Sloan Kettering Cancer Center. All patients were counseled about the benefits of chemoprevention, yet only 2 (4.7%) decided definitively to accept tamoxifen therapy. Similarly, Vogel et al [54] reported that of risk-eligible women who were evaluated for participation in the NSABP’s current chemoprevention trial to compare tamoxifen and raloxifene, only 21% agreed to randomization. In contrast, a diagnosis of atypical hyperplasia substantially escalates a woman’s interest in chemoprevention therapy or clinical trial participation. Vogel et al [54] found that of risk-eligible women who also had a history of atypia, approximately one third agree to randomization. Morrow et al [55] similarly reported that high-risk women are more likely to accept chemoprevention and physicians are more likely to recommend this option if a diagnosis of atypia has been made. Atypia seems to be a more compelling motivation that empowers high-risk women to make difficult decisions regarding risk-reduction strategies. The strength of atypia in identifying participants for chemoprevention trials motivated the NSABP to allow inclusion of abnormal lavage findings into Gail model risk calculations. For assessment of chemoprevention trial eligibility, atypia that is detected on ductal lavage cytology is entered into the Gail model risk estimate as the equivalent of one biopsy with atypia; normal ductal lavage data are not entered into the risk estimate calculations. Hence, past studies have confirmed that atypical hyperplasia that is detected by FNA biopsy, nipple aspirate, or surgical biopsy is a reliable risk

56

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

factor for breast cancer and a marker of risk that is suppressed readily by chemoprevention with tamoxifen. Recent studies demonstrated that ductal lavage is a well-tolerated, minimally-invasive maneuver that can retrieve cells readily to aide in the detection of atypia. It can be assumed reasonably that atypia that is diagnosed with this modality has similar predictive value as when it is detected by the older methods. What information is available from clinically-active, contemporary ductal lavage programs and how should patients be followed? The next issue to be addressed is related to integration of ductal lavage into contemporary clinical practice, with an appraisal of its availability and what the technology has revealed. The length of time that is necessary to assimilate case series, analyze data, and proceed through peer review to publication easily can exceed several years; a review of abstracts that are presented at major academic meetings provides an intermediate barometer of a new technology’s use. Therefore, it is worthwhile to appraise data that are presented in abstract and manuscript form regarding experiences with the ductal lavage technology since the Dooley et al [52] 2001 multi-center study publication. Interest in ductal lavage has been increasing as reflected by the number of abstracts that were presented at some of the national academic meetings that have focused on applications of this technology. A review of the program proceedings for the meetings of the American Society of Clinical Oncology and the San Antonio Breast Cancer Symposium revealed 5 ductal lavage–related presentations in 2001, 10 in 2002, and 11 in 2003. Reported results of programs that use ductal lavage as a clinical risk assessment adjunct Generally, investigators who reported on individual institution/practice experiences with the ductal lavage procedure had technical success in 70% to 80% of cases, comparable to the multi-center study [52]. This yield may be improved by technical modifications, such as topical 2% nitroglycerine applied to the nipple approximately 30 minutes hour before attempting duct cannulation. Golewale et al [56] found that this increased lavage success rates by more than 20%. As shown in Table 3, data on the prevalence of atypia in ductal lavage specimens are consistent with earlier investigations of atypia prevalence from other tissue sources that were based on the study populations. Atypia is uncommon (2%) in unselected, general populations as demonstrated by studies of nipple aspirates on average-risk, asymptomatic women [27]. Rates of atypia are higher (15%–26%) when detected as coexisting lesions in cancer-bearing biopsy specimens and intermediate (2%–10%) when tissue specimens from noncancerous biopsies are evaluated [35–37,39,40]. As expected, the prevalence of atypia that is detected on ductal lavage of

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

57

Table 3 Prevalence of atypia Tissue source

Study population

Nipple aspirates Random FNA Ductal lavage

General population High-risk women High-risk women, multi-center study [52] High-risk women, various backgrounds 12 patients undergoing surgical biopsy, risk status unknown [57] 33 high-risk patients [58] 24 high-risk patients [59] 90 high-risk patients [60] 77 high-risk patients [61] 93 high-risk patients [62] 30 known or suspected BRCA mutation carriers [63] 24 known or suspected BRCA mutation carriers [64] 138 ductography cases with pathologic nipple discharge [65] 29 cancerous breasts undergoing mastectomy [66] Surgical biopsy cases, benign Surgical biopsy cases, cancerous

Surgical biopsy

Proportion of study population who had atypia 2% 12% 6–17%, minimal to moderate-severe

1/12; 8%

1/24; 4.2% 4/90; 4.5% 19/77; 25% 25/93; 27% 7/30; 23% 13/24; 54% 112/138; 81% 10/29; 34% 2–11% 14–26%

high-risk women tends to fall into the intermediate category but varies with categories of risk and the type of patient who is evaluated. Ductal lavage that is performed in the accepted clinical fashion, as a risk assessment adjunct, reveals cytologic atypia in 4% to 27% of ‘‘conventional’’ high-risk cases; prevalence of atypia increases to 23% to 54% in known or suspected carriers of BRCA mutation who undergo lavage. Studies of ductal lavage in the investigational setting revealed even higher rates of atypia among women who underwent lavage in conjunction with ductography that was performed for a pathologic nipple discharge (81%) and in the cancerous breasts of patients who underwent mastectomy (34%) (see Table 3). Dooley et al [61] reported that ductal lavage findings are influential in guiding decisions regarding tamoxifen among women who are high-risk on the basis of a personal history of unilateral breast cancer. They found that ductal lavage cytology in the contralateral breast led to a 16% increase in the use of tamoxifen. Ductal lavage also has been incorporated into comprehensive screening programs for women who have a hereditary risk of breast cancer, where it may be a useful adjunct to innovative surveillance modalities, such as magnetic resonance imaging (MRI) in these carefullyselected patients [63,67,68]. Other categories of risk that would be interesting to study in ductal lavage programs include women who have

58

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

a history of therapeutic chest wall irradiation (eg, Hodgkin’s disease) or a history of prolonged postmenopausal hormone replacement therapy. Correlation of ductal lavage cytology with surgical histopathology Ductal lavage catheters are approved by the Food and Drug Administration for use in performing galactography; at least two studies that correlated diagnostic ductogram findings to lavage results were reported [65,69]. These studies showed that the lavage procedure can be complementary to the ductogram; however, its sensitivity is not sufficiently high that a normal lavage cytology can eliminate the need for surgical follow-up if otherwise indicated by the clinical scenario and ductographic imaging. Lawler et al [65] conducted ductal lavage in the form of ductogram flush washings in 138 patients who were symptomatic for a pathologic nipple discharge and who subsequently underwent surgical biopsy. Atypical cells were found in 112 (81%) of the lavage specimens; follow-up biopsies revealed lobular carcinoma in situ, atypical ductal hyperplasia (ADH), and papillomas in nearly 90% of these cases. The lavage cytology was nonmalignant in 24 of 27 cases that were found to be cancerous at surgical biopsy (benign cytology in 3 of 27 [11%] and atypical in 21 of 27 [78%]). Although atypia on ductal lavage cytology reasonably may be assumed to represent a marker of increased risk, studies that correlated ductal lavage with pathology from cancerous mastectomy specimens and from surgical biopsy specimens confirmed that the lavaged ductal system will not reflect a documented site of disease consistently in the studied breast. These reports reaffirm that ductal lavage is not a cancer screening or detection procedure. Brogi et al [66] performed ductal lavage in 26 breasts that underwent mastectomy for cancer; none of the retrieved cytologies was clearly malignant. Khan et al [70] reported an innovative project that involved ductal lavage of cancerous breasts that was accompanied by the creation of a castlike impression of the lavaged ductal system by injection of a gelatinous material after the mastectomy had been completed. This topographic study demonstrated a correlation between the location of the cancer and the lavaged ductal system in only two thirds of cases. Gabram et al [57] performed ductal lavage on 12 women before diagnostic open biopsy; they found discordant results between the lavage cytology and the surgical histopathology in nearly half of the cases. At best, the procedure only evaluates cells from a portion of the ductal system that has been cannulated. Therefore, it is unlikely to provide sufficient information regarding the entire breast. Dooley et al [71] performed ductal lavage in the contralateral breasts of women who underwent breast cancer surgery and detected atypia in 32%, 22%, and 6.7% of T1a/b, T1c, and T2 lesions, respectively. Early pathology studies of atypia demonstrated that breast cancer risk tends to return to baseline after approximately 5 years if no other risk-related events intervene.

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

59

If a cancer arises within a field of high-risk breast tissue, then surrounding (including contralateral) breast tissue that previously harbored atypical hyperplasia may continue to undergo regression in accordance with the data indicating a return to baseline risk 5 years after the atypia was detected. The cancerous lesion may lose an association progressively with atypia over time and will not remain necessarily associated with a fluid-yielding ductal system. Furthermore, breast cancer pathogenesis is likely to be heterogeneous; not every area of ductal atypia is committed to progressing to invasive cancer if left untreated. Conversely, some cancers may arise without having passed through an atypical hyperplasia precursor phase [22]. Potential limitations to the incorporation of ductal lavage into breast cancer risk assessment programs One issue regarding ductal lavage that requires further investigation is related to the reproducibility of resulting cytologic analyses. Dr. Bonnie King participated in the cytopathology review for the Dooley et al [52] study of ductal lavage and the Wrensch et al [27] study of nipple aspirates and provided inferential evidence that atypical cells that are retrieved by both interventions can be well-standardized. Nonetheless, there is established precedent to confirm the subjectivity and variation that can exist in histopathologic assessment of borderline breast lesions [72] and one might infer reasonably that differences in the cytologic evaluation of ductal lavage specimens also might exist. Therefore, it is important to evaluate the interlaboratory reproducibility of ductal lavage analyses. There is no uniformity in reimbursement policies that are offered by third party payers to cover expenses that are generated by the ductal lavage procedure. This poses a significant limitation to the availability of ductal lavage as a clinical service, even among appropriately-selected high-risk patients. Ozanne and Esserman constructed a mathematical cost-effectiveness model that demonstrated that ductal lavage could lead to health care cost-savings if the expected rates of atypia detection resulted in a decreased number of breast cancers that required treatment with tamoxifen [72a]. Patients who are considering the ductal lavage procedure for risk assessment must understand that significant out-of-pocket expenses may be incurred. Access to the ductal lavage technology clearly is not universal. Follow-up management strategies in ductal lavage cases After a decision has been made to perform ductal lavage as a risk assessment strategy, the clinician must be prepared to handle the results and offer appropriate postprocedural counseling. An essential concept for the patient and health care provider to understand is that normal cytology on ductal lavage does not alter the magnitude of any pre-existing breast cancer risk factors; it does not replace the need for routine breast cancer

60

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

surveillance nor does it have any implications regarding the possible presence or absence of an occult malignancy. The procedure is clinically significant only if atypia is detected, in which case it is relevant in strengthening the patient’s risk profile and may influence decisions regarding risk-reducing strategies. Morrow et al [73] developed a comprehensive schema that described the integration of ductal lavage into the breast cancer risk profile; an algorithm for handling abnormal lavage results is presented in Table 4. Benign cytology should prompt reinforcement of the routine breast cancer surveillance message by way of self-examination, clinical examination, and screening mammography. When atypia is detected, a second cytopathology opinion is reasonable, but not absolutely necessary. Quantified risk should be recalculated by accounting for the atypical lavage in a Gail model risk estimate as the equivalent of one breast biopsy with atypia. Counseling regarding riskreduction strategies should follow, including a reassessment of eligibility for chemoprevention protocols. Follow-up lavage at 6 to 12 months later, regardless of whether the patient opts to take tamoxifen, is a reasonable prospect but is of uncertain significance because the yield of sequential procedures has not been studied completely. One potential concern might be the possibility of variation in lavage cytology that is related to the menstrual cycle. Although serial lavage results have not been reported on any large patient cohorts, extrapolation from the literature on direct nipple aspirates offers some encouraging insight that ductal fluid findings are independent of the menstrual cycle. Mitchell et al [74] performed weekly nipple aspirates on 15 premenopausal volunteers and no significant differences in cellular profile were detected over the course of two menstrual cycles. A diagnostic dilemma is created in the disturbing circumstance of frankly cancerous cells that are detected within a ductal lavage specimen. Because the ductal lavage procedure is only appropriate as a risk assessment adjunct for a woman who does not have any evidence of a pre-existing cancer, this scenario is rare (occurring in fewer than 1% of the cases that were reported in the multi-center study [52]) and presumably would occur in cases that are associated with a negative mammogram and clinical examination. Nonetheless, it is worthwhile to review and repeat the clinical evaluation of the affected breast in these cases. Confirmation of the lavage findings by second opinion cytopathology also is useful. Malignant-appearing cytology can be caused by a variety of benign and cancerous primary breast lesions, including atypical hyperplasia and papillomatosis. A target for biopsy and histopathologic correlation should be sought aggressively with maneuvers, such as ductography (perhaps performed in conjunction with repeat lavage), whole-breast ultrasound, breast MRI, and ductoscopy. If all studies are negative for identifying a source of the cancerous cytology, the patient is offered options of observation versus chemoprevention, with consideration of complete re-evaluation (including a repeat lavage) 6 to 12 months later. The optimal interval between sequential ductal lavage procedures has not

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

61

Table 4 Management of patients who have abnormal lavage Lavage results Management

Atypical cytology

Malignant cytology

Patient counseling

Recalculate Gail model risk estimate, with lavage counted as one biopsy with atypia Repeat risk reduction counseling, including consideration of chemoprevention trial participation

Possible etiologies: Occult malignancy ADH Papillary lesion(s)

Confirmation of results by second review

Optional; not necessary

Necessary; options include: Review of cytopathology Consider repeat lavage

Work-up to identify target lesion for biopsy

Not necessary if prelavage breast imaging and examination were normal

Necessary: Repeat breast examination Review all prelavage imaging Repeat breast imaging, including diagnostic mammography views and breast ultrasound Ductography Breast MRI Consider ductoscopy, if available

Management options

Chemoprevention trial if Observation and eligible re-evaluation in Observation and re-evaluation 3–6 months in 6–12 mo Tamoxifen and re-evaluation Tamoxifen and re-evaluation in 3–6 months in 6–12 mo Prophylactic mastectomy Prophylactic mastectomy only to be considered by the highest-risk women (eg BRCA mutation carriers)

Abbreviation: ADH, atypical ductal hyperplasia.

been defined. The option of performing a blind terminal duct excision might be considered; however, the disadvantage of this approach is that a tumor that is shedding malignant cells into the ductal system could be located peripheral to the subareolar ductal system, and therefore, could easily be missed with this surgical procedure. Furthermore, after the terminal duct apparatus has been divided and resected, the option of repeat cannulation

62

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

and ductal lavage is lost. Patients who undergo prophylactic mastectomy because of cancerous lavage cytology should be forewarned of the possibility that identification of a primary tumor mass in the breast may be difficult, if not impossible. How can the ductal lavage technology be used in the research setting? Presentations at national meetings provide evidence of active research endeavors; the proportion of ductal lavage–related projects that have been related to basic scientific or translational research also has increased steadily over the past few years. In 2001 and 2002, nearly all such work was related to clinical applications of the ductal lavage procedure. During 2003 and 2004, however, more than half of the reported studies featured ductal lavage as a research tool. Investigators have published successful results in detecting molecular markers, such as Her2/neu [75], basic fibroblastic growth factor [76], and carcinoembryonic factor [77], in nipple aspirate fluid. Ductal lavage may be an additional means of monitoring expression of these proteins and has been used to detect evidence of cancer by methylation-specific polymerase chain reaction (PCR) [78]. The evolution of microarray technology for the analysis of DNA content and ‘‘genetic profiling’’ has added another layer to the

Table 5 Use of ductal lavage as a research tool Study

Technology applied to lavage specimen

Evron et al, 2001 [78]

Methylation-specific PCR

Kim et al, 2002 [86]

Fluorescence in situ hybridization Karyometric measurements

Walling et al, 2003 [87] Fournier et al, 2003 [88]

Surface enhanced laser desorption/ionization protein chip mass spectrometry Misell et al, 2003 [89] Stable isotope mass spectrometry Chatterton et al, 2003 [90] Immunohistochemistry Arun et al, 2003 [84]

Features evaluated Promoter hypermethylation of cyclin D2, RAR-b2, Twist, and ER (a cancer-specific phenomenon) Aneusomy at chromosomes 1, 8, 11, and 17 Nuclear chromatin characteristics Proteomic patterns

Breast epithelial cell proliferation Progesterone, EGF, cathepsin D, estrone sulfate Cytology, immunohistochemistry Changes in cytology; levels pre- and post celecoxib of Ki-67; COX-2; EGFR; therapy for chemoprevention and p53

Abbreviations: COX, cyclooxygenase; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; ER, estrogen receptor; RAR, retinoic acid receptor; ROBE, routine operative breast endoscopy.

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

Box 1. Issues that are related to the use of ductal lavage Documented, established issues The magnitude of the breast cancer burden among American women is substantial Breast cancer risk-reduction strategies are available, but are associated with potential morbidity from adverse effects Existing risk assessment methods have limitations and alternative, individualized risk-assessment strategies are needed Atypia consistently and reliably identifies one category of highrisk women and is particularly sensitive to the antiproliferative effects of chemoprevention agents (eg, tamoxifen) Atypia can be detected on FNAs, core biopsy specimens, tissue specimens, nipple aspirates, and ductal lavage cytology Ductal lavage specimens are more likely to yield cytologicallyevaluable fluid compared with direct nipple aspirates Ductal lavage is not a breast cancer detection/screening modality Issues that have not been documented Does atypia within ductal lavage specimens confer the same magnitude of future breast cancer risk compared with atypia detected via other sources? Does atypia on ductal lavage specimens magnify the likelihood of breast cancer development in women who already are deemed high-risk on the basis of contralateral breast cancer or BRCA mutation? Is the cytologic interpretation of ductal lavage reproducible or is there interlaboratory variation? Can ductal lavage correct the deficiencies of existing risk assessment methods and improve the individualized risk assessment of nonwhite American women and women who were exposed to chest wall irradiation or prolonged postmenopausal hormone replacement therapy? Is ductal lavage cost effective? Is the lavaged ductal system likely to be the highest-risk area of the breast? What is the significance of cancer cells that are detected in lavage fluid? What is the significance of serial ductal lavage studies? Will cytology from lavage specimens change over time and in accordance with risk-reducing interventions? Are there other applications for ductal lavage (eg, the detection of risk-associated proteins or in providing cells that may be studied with microarray technology?

63

64

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

sophistication and complexity of tissue research [79]. The prospect of applying these strategies to cells that are retrieved from ductal lavage is provocative. It is important to consider the influence of nonepithelial cytology components of lavage fluid on these investigations. In particular, the substantial contribution of macrophage-derived mammary foam cells to the cellular content of nipple aspirates, as well as lavage fluid was documented by Krishnamurthy et al [80] and King et al [81]. These foam cells may affect the observed microarray and DNA patterns. Petricoin et al [82] also explored the concept of using proteomic pattern diagnostics on ductal lavage fluid as a means of identifying biomarkers of breast cancer risk. The primary purpose of the ductal lavage procedure in the clinical setting is to provide a source for cytology evaluation. Therefore, it is essential that adequacy of the lavage cellularity for achieving this goal is preserved. As additional analytic technologies are applied to the retrieved sample, strategies for safely dividing and allocating the lavage aspirate must be tested. Accordingly, Clark et al [83] analyzed and confirmed the yield from ductal lavage after the specimen was split at the bedside and found equivalence of the divided specimens in regard to cytologic adequacy. Table 5 summarizes the research strategies that are being explored for application in the setting of ductal lavage specimen. Some of these pilot studies include the use of ductal lavage to monitor response to innovative chemoprevention agents such as celecoxocib [66]; studying expression of various hormonally-active active substances by way of immunohistochemistry; and application of proteomic assays. As reported elsewhere [85], the known and unknown issues that are related to the history and possible future of ductal lavage can be summarized as follows (Box 1).

References [1] Ries L, Eisner M, Kosary M. SEER Cancer Statistics Review, 1973–1999. Bethesda (MD): National Cancer Institute; 2002. [2] Hartmann LC, et al. Efficacy of bilateral prophylactic mastectomy in women with a family history of breast cancer. N Engl J Med 1999;340(2):77–84. [3] Rebbeck TR, et al. Prophylactic oophorectomy in carriers of BRCA1 or BRCA2 mutations. N Engl J Med 2002;346(21):1616–22. [4] Fisher B, et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 1998;90(18):1371–88. [5] Baum M, Budzar AU, Cuzick J, Forbes J, Houghton JH, Klijn JG, et al. Anastrozole alone or in combination with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early breast cancer: first results of the ATAC randomised trial. Lancet 2002;359(9324):2131–9. [6] Claus EB, Risch N, Thompson WD. The calculation of breast cancer risk for women with a first degree family history of ovarian cancer. Breast Cancer Res Treat 1993;28(2):115–20. [7] Kelsey J, Gammon M. The epidemiology of breast cancer. CA Cancer J Clin 1991;41(3): 146–65.

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

65

[8] Colditz GA, Rosner B. Cumulative risk of breast cancer to age 70 years according to risk factor status: data from the Nurses’ Health Study. Am J Epidemiol 2000;152(10):950–64. [9] Rossouw JE, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 2002;288(3):321–33. [10] Endogenous Hormones and Breast Cancer Collaborative Group. Body mass index, serum sex hormones, and breast cancer risk in postmenopausal women. J Natl Cancer Inst 2003;95: 1218–26. [11] Wolfe JN. Risk for breast cancer development determined by mammographic parenchymal pattern. Cancer 1976;37(5):2486–92. [12] Cuzick J. Epidemiology of breast cancer: selected highlights. Breast 2003;12:405–11. [13] Collaborative Group on Hormonal Factors in Breast Cancer. Familial breast cancer: collaborative reanalysis of individual data from 52 epidemiological studies including 58,209 women with breast cancer and 101,986 women without the disease. Lancet 2001;358:1389–99. [14] Gail MH, et al. Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. J Natl Cancer Inst 1989;81(24):1879–86. [15] McTiernan A, et al. Comparisons of two breast cancer risk estimates in women with a family history of breast cancer. Cancer Epidemiol Biomarkers Prev 2001;10(4):333–8. [16] Bondy ML, et al. Validation of a breast cancer risk assessment model in women with a positive family history. J Natl Cancer Inst 1994;86(8):620–5. [17] Costantino JP, et al. Validation studies for models projecting the risk of invasive and total breast cancer incidence. J Natl Cancer Inst 1999;91(18):1541–8. [18] Spiegelman D, et al. Validation of the Gail et al., model for predicting individual breast cancer risk. J Natl Cancer Inst 1994;86(8):600–7. [19] Rockhill B, et al. Validation of the Gail et al., model of breast cancer risk prediction and implications for chemoprevention. J Natl Cancer Inst 2001;93(5):358–66. [20] Newman LA, Rockhill B, Bondy ML, Abrams JA, Berlin JA, Colditz GA, et al. Validation of the Gail breast cancer risk assessment model in African American women based on a multi-center case-control study of 3,283 African American and 5,974 white American women. Presented at the American Society of Clinical Oncology, 38th Annual Meeting. Orlando, Florida, May 18–21, 2002. [21] Euhus DM. Understanding mathematical models for breast cancer risk assessment and counseling. Breast J 2001;7(4):224–32. [22] Singletary SE. A working model for the time sequence of genetic changes in breast tumorigenesis. J Am Coll Surg 2002;194(2):202–16. [23] Bhathal PS, et al. Frequency of benign and malignant breast lesions in 207 consecutive autopsies in Australian women. Br J Cancer 1985;51(2):271–8. [24] Nielsen M, et al. Breast cancer and atypia among young and middle-aged women: a study of 110 medicolegal autopsies. Br J Cancer 1987;56(6):814–9. [25] Hutchinson WB, et al. Risk of breast cancer in women with benign breast disease. J Natl Cancer Inst 1980;65(1):13–20. [26] Dupont WD, Page DL. Risk factors for breast cancer in women with proliferative breast disease. N Engl J Med 1985;312(3):146–51. [27] Wrensch MR, et al. Breast cancer incidence in women with abnormal cytology in nipple aspirates of breast fluid. Am J Epidemiol 1992;135(2):130–41. [28] Page DL, Dupont WD. Anatomic indicators (histologic and cytologic) of increased breast cancer risk. Breast Cancer Res Treat 1993;28(2):157–66. [29] Fabian CJ, et al. Short-term breast cancer prediction by random periareolar fine-needle aspiration cytology and the Gail risk model. J Natl Cancer Inst 2000;92(15):1217–27. [30] Wrensch MR, et al. Breast cancer risk in women with abnormal cytology in nipple aspirates of breast fluid. J Natl Cancer Inst 2001;93(23):1791–8. [31] Dupont WD, et al. Estrogen replacement therapy in women with a history of proliferative breast disease. Cancer 1999;85(6):1277–83.

66

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

[32] Bodian CA. Benign breast diseases, carcinoma in situ, and breast cancer risk. Epidemiol Rev 1993;15(1):177–87. [33] Bodian CA, et al. Prognostic significance of benign proliferative breast disease. Cancer 1993; 71(12):3896–907. [34] Carter CL, et al. A prospective study of the development of breast cancer in 16,692 women with benign breast disease. Am J Epidemiol 1988;128(3):467–77. [35] London SJ, et al. A prospective study of benign breast disease and the risk of breast cancer. JAMA 1992;267(7):941–4. [36] McDivitt RW, et al. Histologic types of benign breast disease and the risk for breast cancer. The Cancer and Steroid Hormone Study Group. Cancer 1992;69(6):1408–14. [37] Palli D, et al. Benign breast disease and breast cancer: a case-control study in a cohort in Italy. Int J Cancer 1991;47(5):703–6. [38] Krieger N, Hiatt RA. Risk of breast cancer after benign breast diseases. Variation by histologic type, degree of atypia, age at biopsy, and length of follow-up. Am J Epidemiol 1992;135(6):619–31. [39] Dupont WD, et al. Breast cancer risk associated with proliferative breast disease and atypical hyperplasia. Cancer 1993;71(4):1258–65. [40] Byrne C, et al. Biopsy confirmed benign breast disease, postmenopausal use of exogenous female hormones, and breast carcinoma risk. Cancer 2000;89(10):2046–52. [41] Carpenter C, Love SM. Ductal cells, hyperplasia and breast cancer risk: results from a longterm follow-up study of lavage patients. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2002. [42] Marshall CJ, et al. Cytologic identification of clinically occult proliferative breast disease in women with a family history of breast cancer. Am J Clin Pathol 1991;95(2):157–65. [43] Skolnick MH, et al. Inheritance of proliferative breast disease in breast cancer kindreds. Science 1990;250(4988):1715–20. [44] Ward JH, et al. Detection of proliferative breast disease by four-quadrant, fine-needle aspiration. J Natl Cancer Inst 1990;82(11):964–6. [45] Morrow M, Jordan V. Managing breast cancer risk. In: Jordan V, editor. London: Decker Inc.; 2003. p. 153. [46] Stoler D, Stewart C, Stomper P. Breast epithelium procurement from stereotactic core biopsy washings: flow cytometry-sorted cell count analysis. Clin Cancer Res 2002;8:428–32. [47] Papanicolaou GN, et al. Cytologic evaluation of breast secretions. Ann N Y Acad Sci 1956; 63(6):1409–21. [48] Papanicolaou GN, et al. Exfoliative cytology of the human mammary gland and its value in the diagnosis of cancer and other diseases of the breast. Cancer 1958;11(2):377–409. [49] Sartorius OW, et al. Cytologic evaluation of breast fluid in the detection of breast disease. J Natl Cancer Inst 1977;59(4):1073–80. [50] Newman LA, Blake C. Ductal lavage for breast cancer risk assessment. Cancer Control 2002;9(6):473–9. [51] Tchou J, Hou R, Jordan RAV, Morrow M. Patient acceptance of tamoxifen as chemoprevention, abstract 2216. Lynn Sagar Breast Cancer Symposium, Chicago, IL, September 21, 2003. [52] Dooley WC, et al. Ductal lavage for detection of cellular atypia in women at high risk for breast cancer. J Natl Cancer Inst 2001;93(21):1624–32. [53] Port ER, et al. Patient reluctance toward tamoxifen use for breast cancer primary prevention. Ann Surg Oncol 2001;8(7):580–5. [54] Vogel VG, et al. Re: tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 2002;94(19): 1504. [55] Morrow M, et al. Acceptance of tamoxifen chemoprevention varies with the cause of breast cancer risk. Presented at the 2003 American Society of Clinical Oncology Annual Symposium. Chicago, 2003.

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

67

[56] Golewale N, et al. Technical modifications of ductal lavage to improve cell yield. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [57] Gabram S, et al. Correlation of nipple aspiration and ductal lavage with histopathologic findings for patients prior to scheduled breast biopsy. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [58] Ganz PA, et al. Identification of premalignant and malignant breast cells in mammogramand physical exam-negative women by ductal lavage: results from a multicenter trial. Presented at the American Society of Clinical Oncology Annual Symposium. 2001. [59] Masood S, et al. Exfoliative breast cytopathology: an experience with ductal lavage. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2002. [60] Willey S, et al. Ductal lavage: initial 18 month experience at a single institution. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2002. [61] Dooley WC, Clarke A. Impact of ductal lavage on risk management in a breast surgical oncology practice. Presented at the American Society of Clinical Oncology Annual Symposium. Orlando, 2003. [62] Cazzaniga M, et al. Identification of atypical cells by ductal lavage in women at increased high risk for breast cancer. Presented at the American Society of Clinical Oncology Annual Symposium. 2001. [63] Hartman AR, et al. Breast magnetic resonance image screening and ductal lavage in women at high genetic risk for breast carcinoma. Cancer 2004;100(3):479–89. [64] Mitchell G, et al. Using nipple aspiration and ductal lavage to assess the ductal epithelium of BRCA 1/2 mutation-carriers. Presented at the American Society of Clinical Oncology Annual Symposium. Orlando, 2002. [65] Lawler M, et al. Cytologic evaluation of duct washings does not reliably detect lactiferous duct pathology. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2001. [66] Brogi E, et al. Ductal lavage in patients undergoing mastectomy for mammary carcinoma: a correlative study. Cancer 2003;98(10):2170–6. [67] Hartman AR, et al. Comprehensive screening using breast MRI and ductal lavage in highrisk women. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2002. [68] Hartman AR, et al. Results from a pilot breast cancer screening trial using a combination of clinical breast exam, mammography, breast MRI, and ductal lavage in a high-risk population. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [69] Rossmann M, et al. Galactography versus ductal lavage. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2001. [70] Khan SA, et al. Ductal lavage findings in women with known breast cancer undergoing mastectomy. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2002. [71] Dooley WC, Clark A, Parker J. The frequency of ductal atypia in high-risk breast tissue. Presented at the 56th Annual Cancer Symposium for the Society of Surgical Oncology. Los Angeles, 2003. [72] Rosai J. Borderline epithelial lesions of the breast. Am J Surg Pathol 1991;15(3):209–21. [72a] Ozanne EM, Esserman LT. Cost effectiveness of ductal lavage: a breast cancer risk assessment technique. Abstract 547. 2002 San Antonio Breast Cancer Symposium, San Antonio, Texas, December 13, 2002. [73] Morrow M, et al. Evaluation and management of the woman with an abnormal ductal lavage. J Am Coll Surg 2002;194(5):648–56. [74] Mitchell G, et al. Cellular characteristics of nipple aspiration fluid during the menstrual cycle in healthy premenopausal women. Cytopathology 2001;12(3):184–96. [75] Kuerer HM, et al. High and differential expression of HER-2/neu extracellular domain in bilateral ductal fluids from women with unilateral invasive breast cancer. Clin Cancer Res 2003;9(2):601–5. [76] Liu Y, et al. Breast-cancer diagnosis with nipple fluid bFGF. Lancet 2000;356(9229):567.

68

A. Rivers, L.A. Newman / Surg Oncol Clin N Am 14 (2005) 45–68

[77] Zhao Y, et al. Nipple fluid carcinoembryonic antigen and prostate-specific antigen in cancerbearing and tumor-free breasts. J Clin Oncol 2001;19(5):1462–7. [78] Evron E, et al. Detection of breast cancer cells in ductal lavage fluid by methylation-specific PCR. Lancet 2001;357(9265):1335–6. [79] Weber BL. Cancer genomics. Cancer Cell 2002;1(1):37–47. [80] Krishnamurthy S, et al. Characterization of foam cells in nipple aspirate fluid. Diagn Cytopathol 2002;27(5):261–4 [discussion 265]. [81] King BL, et al. Immunocytochemical analysis of breast cells obtained by ductal lavage. Cancer 2002;96(4):244–9. [82] Petricoin EE, Paweletz CP, Liotta LA. Clinical applications of proteomics: proteomic pattern diagnostics. J Mammary Gland Biol Neoplasia 2002;7(4):433–40. [83] Clark P, et al. Ductal lavage specimens can be successfully split at the bedside without compromising cytologic diagnosis. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [84] Arun B, et al. Phase II chemoprevention of celecoxib using ductal lavage. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [85] Newman LA. Ductal lavage: what we know and what we don’t. Oncology 2004;18(2): 179–85. [86] Kim J, et al. Molecular characterization of mammary ductal epithelial cells using interphase fluorescence in-situ hybridization (FISH). Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2002. [87] Walling E, et al. Nuclear chromatin characteristics of breast epithelial cells obtained by ductal lavage. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [88] Fournier K, et al. Feasibility of proteomic evaluation of ductal lavage fluid for breast diseases. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [89] Misell L, et al. A stable-isotope mass spectrometric method for measuring breast epithelial cell proliferation in vivo in women: a promising biomarker for prevention studies. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003. [90] Chatterton R, Geiger A, Khan SA. Influence of ovarian secretions, oral contraceptives, and hormone replacement therapy on progesterone concentrations in breast fluids. Presented at the San Antonio Breast Cancer Symposium. San Antonio, 2003.