The success of the National SurgicalAdjuvant Breast and BowelProject (NSABP) trial P-01at showing that we now have an effectivemeans to prevent breast cancerposes larger and more seriousquestions: Who should receivechemoprevention, and at what pointin life should this occur? The designof the P-01 study allowed many womento enroll who, according to Gailmodel calculations, were at a less than 1% per year risk of subsequent breastcancer during the expected 5-yeartreatment period. These lower-risk individualsseemed to have less benefitthan those patients at much higherrisk. Other similar prevention studiesseem to confirm this observation.
The success of the National SurgicalAdjuvant Breast and BowelProject (NSABP) trial P-01at showing that we now have an effectivemeans to prevent breast cancerposes larger and more seriousquestions: Who should receivechemoprevention, and at what pointin life should this occur? The designof the P-01 study allowed many womento enroll who, according to Gailmodel calculations, were at a less than 1% per year risk of subsequent breastcancer during the expected 5-yeartreatment period. These lower-risk individualsseemed to have less benefitthan those patients at much higherrisk. Other similar prevention studiesseem to confirm this observation.The side effects of current preventiondrugs are significant enough thatmany patients and their doctors stilldo not routinely consider chemopreventionof breast cancer as importantas the prevention of heart attack andstroke by treating elevated blood pressure.In the case of hypertension, weknow that there are certain levels atwhich the relative benefit can be seenquickly when antihypertensives are started. When we review the NSABPP-01 data, it is clear that those at highestrisk-patients with a prior diagnosisof atypical hyperplasia-derivethe greatest benefit from selective estrogen-receptor modulator (SERM)therapy.Need for Atypia Screening
Just as internists have a sphygmomanometerto screen for hypertension,we need a simple way to screen foratypical hyperplasia. This is where allthe problems start, as neither physicalexam nor mammography can routinelyclaim to detect atypical hyperplasias,except by chance. Even the emerging imaging technologies, suchas magnetic resonance imaging (MRI)and nuclear imaging, fail to detectatypia. Only the old technologiesof nipple aspiration, as shown byWrensch and Petrakis,[1,2] and randompooled fine-needle aspiration, asshown by Fabian and Kimler, haveshown any ability to detect atypia.Neither has been shown to detect allatypia, but each (when atypia is found)is associated with a fivefold relativerisk for malignancy developmentin the next 10 years. These dataclosely parallel the histopathologicstudies of Dupont and Page, whichhave defined our thinking on atypicalhyperplasia.Current Understandingof Ductal Lavage
A new contender as a screeningtest for atypical hyperplasia, ductallavage fares much better at atypia detectioncompared to nipple aspiration.Lavage detects only gross epithelialevents that shed numerous cells intraluminallywithout obstructing theducts. As soon as the ducts becomeobstructed, the ability to screen forabnormality rapidly fades; this can beseen in several reports of its use ininvasive cancers.In her article, Dr. Lisa Newmancorrectly identifies the known andunknown aspects of our currentknowledge about ductal lavage. Thenaysayers will point to the lack of 10-to 20-year follow-up data of lavagedpatients and say we don't know howto interpret the results. We now have20-year follow-up data from both histologicand cytologic studies showingthe equivalence of ductal epithelialatypia in risk stratification of womenfor breast cancer development. If wewere to ignore this wealth of data, wewould need some rational explanationof why lavaged exfoliated ductalcells represent a biased and differentpopulation than those seen histologicallyor collected cytologically with nipple aspirate fluid or fine-needleaspiration.Ductal lavage is clearly enormouslypowerful for allowing repeated accessto abnormal ductal cells for avariety of molecular and proteinomicinvestigations. Furthermore, the abilityto follow patients on chemopreventionagents and detect biologicchanges prior to the frank evolutionof a clinical cancer is extremely important.The real burning question isnot how do we use this in the laboratoryto aid our understanding of theevolution and prevention of breastcancer, but how do we use this toolclinically while we are waiting forthe science to catch up to clinicalrealities?The original lavage study showsthat 24% of "high-risk" individualsshed atypical cells intraluminally,which can be identified. This subsetis likely at greater risk, if we are tobelieve the 20 years of prior data, andvery likely to show the greatest benefitfrom chemoprevention attempts.It seems obvious that if patients withlavaged atypia were started onSERMs, it would not take long for alarge cooperative study to demonstrateeffectiveness at reducing breastcancer incidence. Atypia is commonin the contralateral breast of patientsbeing treated for a current breast cancer.Hence, adding contralateral ductallavage to some of our ongoingcooperative trials involving preventiondrugs, and looking for a differentialeffect among women with andwithout ductal lavage atypia at contralateralprevention, would likelyquickly confirm our inference fromthe prior studies.Future Directions
The greatest power of lavage,however, is only barely being investigated-the ability to identify anatomicallywhere in the breast thepremalignant/malignant changes areoccurring before screening MRI,mammography, or physical exam.Imagine directing imaging to a ductalsegment with intraluminal contrast.The whole world of general surgery outside of breast diseases hasbeen revolutionized by endoscopy inthe past 15 years. Now microendoscopesless than 1 mm in diametercan routinely be used to identify lesionsand biopsy them when theymeasure less than 0.1 mm. This is afresh challenge to our pathologistswho have new specimens halfwaybetween cytology and histology. Newanatomy of the breast is being defined,along with an improved understandingof differing types of intraductalproliferation. With these novel technologies,we are destined to identifymany more women with premalignantbreast diseases.Conclusions
As our understanding evolves, perhapswe won't have to wait until cancerdevelops to start treatment. We caninstead try to prevent progression andmonitor our success or failure. Perhapsprevention needs to be directed at individualductal trees, not systemically.For certain, the prevention of breastcancer needs to be almost as commonas the treatment of hypertension to preventmyocardial infarction and cerebrovascularaccident. We have our workcut out for us-defining who needschemoprevention and when, whichdrugs to use, and how to administerthem. We need to make the whole processas intuitive to the next generationof medical students as hypertensionmanagement was for us.
The authors have nosignificant financial interest or other relationshipwith the manufacturers of any productsor providers of any service mentioned in thisarticle.
Wrensch MR, Petrakis NL, King EB, et al:Breast cancer incidence in women with abnormalcytology in nipple aspirates of breast fluid.Am J Epidemiol 135:130-141, 1992.
Wrensch MR, Petrakis NL, Miike R, etal: Breast cancer risk in women with abnormalcytology in nipple aspirates of breast fluid. JNatl Cancer Inst 93:1791-1798, 2001.
Fabian CJ, Kimler BF, Zalles CM, et al:Short-term breast cancer prediction by randomperiareolar fine-needle aspiration cytology andthe Gail risk model. J Natl Cancer Inst92:1217-1227, 2000.
Dupont WD, Page DL: Risk factors forbreast cancer in women with proliferative breastdisease. N Engl J Med 312:146-151, 1985.