Diagnosis and Surgical Treatment of Suspicious Nonpalpable Breast Lesions and Early Breast Cancer

DIAGNOSIS AND SURGICAL ARTO TREATMENT OF SUSPICIOUS SAARELA NONPALPABLE BREAST Department of Surgery LESIONS AND EARLY BREAST CANCER

1999

ARTO SAARELA

DIAGNOSIS AND SURGICAL TREATMENT OF SUSPICIOUS NONPALPABLE BREAST LESIONS AND EARLY BREAST CANCER

Academic Dissertation to be presented with the assent of the Faculty of Medicine, University of Oulu, for public discussion in Auditorium 1 of the University Hospital of Oulu, on September 24th, 1999, at 12 noon.

Manuscript received 28.8.1999 Accepted 2.9.1999

Communicated by Professor Peter Dean Professor Matti Eskelinen

ISBN 951-42-5360-4 (URL: http://herkules.oulu.fi/isbn9514253604/)

ALSO AVAILABLE IN PRINTED FORMAT

ISBN 951-42-5359-0 ISSN 0355-3221

OULU UNIVERSITY LIBRARY OULU 1999 Saarela,ArtoO.Diagnosisandsurgicaltreatmentofsuspiciousnonpalpablebreast lesionsandearlybreastcancer DepartmentofSurgery,UniversityofOulu,FIN-90401Oulu,Finland 1999 Oulu,Finland (Manuscriptreceived28.8.1999)

Abstract

Thepurposesofthepresentresearchweretoevaluate(1)thevalueofultrasonographicallyguided fine-needleaspirationbiopsy(US-FNAB)innonpalpablesuspiciousbreastlesions,(2)thepreoper- ativeuseofmethylenebluestaininginnonpalpablegalactographicallysuspiciousbreastlesions,(3) thedeterminantsofpositivehistologicmarginsandresidualcancerinwire-guidedbiopsy(WGB)of nonpalpablebreastcancerandinlumpectomyforearlybreastcancerandthedeterminantsofposi- tiveradiologicmarginsandthecorrelationbetweenradiologicandhistologicmarginsandresidual diseaseinWGBofnonpalpablebreastcancer,(4)theassessmentoflumpectomymarginsbytouch preparationcytologyinearlybreastcancer,and(5)thecosmeticoutcomeofWGBperformedfor benignbreastlesions. ThesensitivityandspecificityofUS-FNABin90nonpalpablebreastlesionswere84%and 93%,respectively.Preoperativemethylenebluestainingwassuccessfulin22outof30(73%) cases,makingsubsequentselectiveminimalvolumemicrodochectomyeasytoperform.Multivari- ateanalysisof21prospectivelyevaluatedvariableswasdoneafter71WGBsofnonpalpablebreast cancerfollowedby54re-excisions.Largemammographiclesionshadmoreoftenpositiveradio- logicmargins.Multifocality,largepathologicsizeandsuperficialexcisionwererelatedtopositive histologicmarginsandmultifocalitytoresidualdiseaseinre-excisions.Thesensitivityandspecific- ityofspecimenradiographyforpredictinghistologicmarginswere38%and81%andthosefor residualdisease27%and79%,respectively.Thecorrespondingfiguresforhistologicmarginsin predictingresidualdiseasewere85%and59%,respectively.Inaprospectiveseriesof55consecu- tivelumpectomiesforearlybreastcancer,positivehistologicmarginswerefoundmoreofteninthe presenceofintraductalcancerandifthepathologicsizeoftheindextumorwaslarge.Residualdis- easewasfoundin38%ofthecaseswithpositiveandin15%ofthecaseswithnegativehistologic margins.Amultifocalandnonpalpableindextumorpredictedresidualcancerin34re-excision specimens.Thesensitivityandspecificityoftouchpreparationcytologyinpredictinghistologic marginswere38%and85%,respectively.InWGB,theoverallcosmesis6monthsaftersurgery wassatisfactoryin75%ofthe101prospectivelyevaluatedpatientswithbenignprovenlesions. Cosmesiswaspoorerafterdeepexcisionsandcomplications. TheresultsindicatethatUS-FNABisausefultoolinevaluatingnonpalpablesuspiciousbreast lesions.Preoperativemethylenebluestainingcruciallyfacilitatesselectiveminimalvolumemicro- dochectomyinthree-quartersofcases.ToobtainfreemarginsinWGB,mammographicallyand pathologicallylargelesionsshouldberemovedwithwiderexcisionsextendingdowntothefascia. However,radiologicmarginsinWGBandhistologicmarginsbothinWGBandinlumpectomyfor earlybreastcancermaybemisleading.Re-excisionofthebiopsysiteofmultifocaltumorsafter WGBandlumpectomyshouldbeconsidered.Thisisalsoimportantaftersuperficialexcisionin WGBduetotheconsiderableriskofresidualdisease.Touchpreparationcytologycannotberecom- mendedfortheassessmentofmarginsinlumpectomyspecimensofearlybreastcancer.Cosmetic outcomeafterWGBofbenignbreastlesionsissatisfactoryin75%ofcases.Deepexcisionsand complicationsendangerthecosmeticoutcome. Preoperativebiopsyandtumorlocalizationmethodshaveproventheirutility;nevertheless,free marginsarestilldifficulttoobtainandtoevaluateaccurately.Thesurgeonmayoftenbeforcedto choosebetweenfreemarginsandanacceptablecosmeticoutcome.

Keywords:Breastneoplasm,wire-guidedbiopsy,margins,cosmesis Aknowledgements

This work was carried out at the Department of Surgery in co-operation with the Depart- ments of Radiology and Pathology, Oulu University Hospital, during the years 1993– 1999. First of all, I owe my sincere gratitude to Professor Matti Kairaluoma, MD, PhD, Head of the Department of Surgery, for support, excellent conditions and working environment at the Department. I wish to express my warmest thanks to my supervisor, Docent Heikki Kiviniemi, MD, PhD, for valuable ideas, advice and trust in me during this research. His enthusiasm sus- tained my research project through disappointments and difficulties. I express my special appreciation to Tarja Rissanen, MD, PhD, for her crucial role in this project. Her valuable advice, criticism and introduction to scientific work extend far beyond the radiological aspects. Docent Kari Haukipuro, MD, PhD, deserves my deepest thanks for all his valuable work on the statistical problems encountered during this work. I am grateful to Professor Matti Eskelinen, MD, PhD, and Professor Peter Dean, MD, PhD, the reviewers, for their constructive advice and criticism of the manuscript. My co-workers, Docent Outi Kaarela, MD, PhD, Kimmo Lähteenmäki, MD, Timo Paloneva, MD and Docent Ylermi Soini, MD, PhD, deserve my warmest thanks. I cordially express my gratitude to the surgeons and personnel of the Department of Surgery for all their extra work and great support. Especially, I thank my colleagues in the Gastroenterological Division and its present head, Docent Seppo Laitinen, MD, PhD. I also thank Jukka Palm, MD, for his valuable support in computor problems. I wish to thank Mrs. Sirkka-Liisa Leinonen for revising the English language of this thesis, the university secretary Mrs. Ritva Huhtala, and the personnel of the Library of the Medical Faculty of Oulu. Finally, my parents Tauno and Anja Saarela deserve my deepest gratitude for their unselfish support and encouragement throughout my life.

Oulu, August 1999 Arto Saarela

Abbreviations

BCS Breast-conserving surgery BCT Breast-conserving surgery and irradiation for early breast cancer DCIS Ductal carcinoma in situ (intraductal carcinoma) EIC Extensive intraductal component FNAB Fine-needle aspiration biopsy US Ultrasonography WGB Wire-guided biopsy

List of original publications

This thesis is based on the following papers, referred to in the text by the Roman nume- rals I-V. I Saarela AO, Kiviniemi HO, Rissanen TJ & Paloneva TK (1996) Nonpalpable breast lesions: pathologic correlation of ultrasonographically guided fine-needle aspiration biopsy. J Ultrasound Med 15: 549-553. II Saarela AO, Kiviniemi HO & Rissanen TJ (1997) Preoperative methylene blue staining of galactographically suspicious breast lesions. Int Surg 82: 403-405. III Saarela AO, Rissanen TJ, Soini Y, Lähteenmäki KM, Haukipuro K, Kaarela O & Kiviniemi HO (1999) Wire-guided biopsy for nonpalpable breast cancer: determinants and correlation of mammographic and histologic margins and residual cancer in re-excisions. Submitted for publication. IV Saarela AO, Paloneva TK, Rissanen TJ & Kiviniemi HO (1997) Determinants of positive histologic margins and residual tumor after lumpectomy for early breast cancer: A prospective study with special reference to touch preparation cytology. J Surg Oncol 66: 248-253. V Saarela AO, Kiviniemi HO, Rissanen TJ, Haukipuro K & Kaarela O (1998) Cos- metic results after wire-guided biopsy of benign breast lesions. J Am Coll Surg 187: 610-615.

Contents

Abstract Aknowledgements Abbreviations List of original publications 1. Introduction...... 13 2. Reviewoftheliterature ...... 15 2.1.Epidemiologyandincidenceofbreastcancer ...... 15 2.2.Diagnosticmethods ...... 15 2.2.1. Noninvasivemethods...... 15 2.2.1.1.Clinicalexamination ...... 15 2.2.1.2.Mammography...... 16 2.2.1.3.Ultrasonography...... 17 2.2.1.4.Othermethods ...... 17 2.2.2. Invasivemethods...... 18 2.2.2.1.Needlebiopsy...... 18 2.2.2.2.Galactography ...... 19 2.2.2.3.Methylenebluestaining...... 20 2.2.2.4.Wire-guidedbiopsy ...... 21 2.3.Treatment...... 23 2.3.1. Breast-conservingsurgery...... 23 2.3.1.1.Lumpectomy...... 23 2.3.1.2.Axillarydissection...... 24 2.3.1.3. Margins, residual disease and local recurrence ...... 24 2.3.1.4. Multifocality and multicentricity ...... 27 2.3.2. Mastectomy...... 27 2.3.3. Cosmesis ...... 28 3. Aimsofthestudy ...... 29 4. Materialsandmethods ...... 30 4.1.Patients...... 30 4.2.Methods ...... 31 4.2.1. US-FNABinnonpalpablebreastlesions(I) ...... 31 4.2.2. Preoperativemethylenebluestaining(II)...... 31 4.2.3. WGBfornonpalpablebreastcancer(III) ...... 31 4.2.4. Marginsandresidualtumorafterlumpectomy(IV) ...... 33 4.2.5. Cosmesis(V)...... 34 4.2.6. Statisticalmethods...... 35 5. Results...... 36 5.1.US-FNABinnonpalpablebrestlesions(I) ...... 36 5.2.Preoperativemethylenebluestaining(II) ...... 38 5.3.WGBfornonpalpablebreastcancer(III)...... 39 5.4.Marginsandresidualtumorafterlumpectomy(IV) ...... 41 5.5.Cosmesis(V) ...... 43 6. Discussion...... 46 6.1.Subjectsandmethods...... 46 6.2.US-FNABinnonpalpablebreastlesions(I)...... 48 6.3.Preoperativemethylenebluestaining(II) ...... 49 6.4.WGBfornonpalpablebreastcancer(III)...... 49 6.5.Marginsandresidualtumorafterlumpectomy(IV) ...... 51 6.6.Cosmesis(V) ...... 52 6.7.Generaldiscussion...... 53 7. Conclusions...... 55 8. References...... 56 1. Introduction

Breast cancer is the most common malignancy in women in Finland (1). The current life- time risk of a woman for breast cancer is approximately 10 % (2,3). Several randomized and controlled population-based mammography screening trials provide consistent evidence that screening mammography decreases breast cancer deaths in women aged 40 to 69 years (4-7). The widespread use of screening mammography coupled with judicious use of needle biopsy and wire-guided biopsy can reduce the rate of breast cancer deaths (8-11) by helping to identify breast cancers that are smaller and at an earlier stage (9,11- 17). It has been predicted that, within a decade, the median maximum diameter of all invasive breast cancers will be 1 cm (14). Among the known prognostic factors, such as age, lymph node involvement and hormone receptor status, tumor size remains an important predictor of both disease-free and overall survival (18-21). It directly correlates with the prognosis, small tumors (<10 mm) having the best prognosis overall (18,21). Early detection of breast cancer is the only known possibility to improve the life expectancy of these patients (2). Approximately 90 % of early nonpalpable breast cancers are curable at this stage (22,23). Women presenting with nonpalpable mammographically detected invasive breast cancer have a favorable prognosis compared to women with palpable breast cancer (24,25). The reported 20-year survival of women with nonpalpable breast cancer has been over 95% (9). The excellent long-term prognosis and low recurrence rate of small T1 breast cancers facilitates modifications and simplification of therapy in the future (14). With the widespread use of screening and clinical mammography for the detection of early breast cancer, the number of suspicious nonpalpable breast lesions requiring a definitive diagnosis has greatly increased. Image-guided needle aspiration biopsies have reduced the need for open biopsy in obtaining reliable tissue samples with lower cost and less morbidity (26,27). Discharge from the nipple is a fairly common reason for remitting women to a surgical evaluation, comprising 3-4.8 % of all causes of remittance (28-30). The sensitivity of mammography for the pathologic processes behind nipple discharge is low (31-33), and although the sensitivity of galactography is high, its specificity is fairly low (34). 14

Despite the advanced image-guided needle biopsies, a wire-guided biopsy (WGB) sample is still often required for a definitive diagnosis of a clinically occult breast lesion. Although approximately 60 % of biopsied lesions are malignant in population screening (35,36), the majority of nonpalpable breast lesions in women outside screening programs have traditionally been benign (37-42), underscoring the importance of the cosmetic outcome of this widely used procedure. In the era of breast-conserving surgery (BCS), a proportion of diagnostic WGBs also prove to be therapeutic after a histologic examination. Breast-conserving surgery and irradiation for early (stage I and II) breast cancer (BCT) is equivalent to modified radical mastectomy in terms of local and regional control as well as overall survival (43-47). The optimal extent of the initial surgical procedure is controversial, but the appropriate minimum requirement is a microscopically uninvolved margin (21). To date, however, no optimal method for assessing microscopic margins has been established. In this study, preoperative ultrasonography (US)-guided fine-needle aspiration biopsy (FNAB) of nonpalpable breast lesions and methylene blue staining in galactographically suspicious ducts were explored. The determinants of positive histologic margins and residual cancer in WGB of nonpalpable breast cancer and in lumpectomy for early breast cancer were studied. Also, the determinants of positive radiologic margins and the correlation between radiologic and histologic margins and residual disease in WGB of nonpalpable breast cancer were studied. Furthermore, the usefulness of touch preparation cytology in the assessment of lumpectomy margins of early breast cancer was evaluated. Finally, the cosmetic outcome and related determinants after WBG of benign lesions were explored. 2. Review of the literature

2.1. Epidemiology and incidence of breast cancer

Breast cancer is overwhelmingly the most common malignancy among women and one of the most actively investigated tumors worldwide. The incidence rates have been highest in northern Europe and North America (7). Age-specific incidence increases clearly up to 45-50 years of age, suggesting that menopause has a protective effect (7,48). The incidence of breast cancer has been increasing steadily during the past decades (48,49), whereas the mortality rates have remained relatively constant for several years (2,7). In Finland, the incidence of breast cancer (the annual number of new cases diagnosed per 100.000 women) was 76 in 1995. In the same year, 3,125 new breast cancer cases were diagnosed and 838 women died of this disease. For comparison, the corresponding figures for colon cancer in women, the second leading primary malignancy, were 691 and 360, respectively.(1)

2.2. Diagnostic methods

2.2.1. Noninvasive methods

2.2.1.1. Clinical examination

Clinically, apart from mammography, the first and most important way of detecting breast cancer is still to find a new palpable lump by self-examination or physical examination of the breast (50). The true positive rate of physical examinations in premammographic screening was 56% according to Gilbertsen et al. (51). The sensitivity of physical examination has been reported to be more than twofold in younger versus older women (52,53) and to decrease substantially in smaller (<1cm) breast cancers (54). Notably, physical examination may discover interval cancers and mammographically negative cancers (55). The impact of breast self-examination on clinical stage and mortality from breast cancer 16 is still controversial (56-58). The combination of physical examination and mammography is superior to either modality alone (55,59), detecting more than 90% of all breast cancers (60).

2.2.1.2. Mammography

Improvements in mammographic techniques since 1960, when Egan (61) described a reproducible technique for x-ray examination of the breast, have resulted in earlier diagnosis and treatment of breast cancer (12,13). Nowadays, an increasing number of nonpalpable breast lesions are found in attempts to achieve earlier cancer diagnosis in the daily clinical practice and in mammographic screening programs. The sensitivity of high-quality mammography for cancer is high (5). A malignant breast tumor is seen at mammography as a stellate, ill-defined or lobulated circumscribed density, or as a group of microcalcifications without a soft tissue tumor (62). Architectural distorsion and asymmetric densities are less common signs of malignancy. It is, however, not always possible to differentiate between malignant and benign lesions based solely on their mammographic appearance. The positive predictive value of mammography varies in different centres and according to the grade of suspicion, being best when associated with population screening (35,36). In mammographically suspicious cases the probability of malignancy averages 10% to 30 %, and in cases highly suggestive of malignancy the yield is greater than 80 %. The frequency of cancer among mammographically probably benign lesions is low, 0.5 % to 1.7 %. (63,64) About half of the women referred to a surgeon for an abnormal mammogram can be spared biopsy in cases where the referring radiologist has not performed a full preoperative diagnostic workup (65,66). The well recognized role of mammography in the detection of breast cancer has been constantly plagued by reports of 3-34 % false-negative rates, the majority of the rates being between 8 and 25 % (61,67-78). The sensitivity of mammography increases with age (73-75), but this has not been confirmed in all studies (77). The dense breasts of young nulliparous women are usually most difficult to evaluate in mammograms, but the use of hormone replacement therapy preserves the pre-menopausal mammographic density postmenopausally for as long as the treatment is used (79). It has been claimed that this therapy decreases the sensitivity of screening mammography (80). The cancers not visualized even in a mammogram review account for 30-52 % of all mammographically negative cancers (67,69-71). One to two percent of all breast cancers present as nipple discharge (81). Even nowadays, the sensitivity of mammography for the pathologic processes behind nipple discharge is low (31-34). In the future, technical improvements, including the digital computer-assisted technique, digital tomosynthesis, digital angiog- raphy, synchrotron light, and laser mammography, may be used to increase the sensitivity and specificity of mammography (82). Nevertheless, conventional mammography remains the most cost-effective tool for breast cancer diagnosis due to its high sensitivity, good availability and low cost. Notably, mammographically negative breast cancer has no negative impact on patient eligibility for breast-conserving surgery (83). 17

Mammography has some important limitations in selecting patients for BCT. Conven- tional mammography underestimates the extent of ductal carcinoma in situ (DCIS) (84), which is detected solely by mammographic microcalcifications in approximately 75% of cases (85). In nonpalpable mammographic abnormalities, microcalcifications constitute one half of the indications for WGB (86). Magnification mammography can improve the correlation between radiologic and histologic tumor size and aid in patient selection (87).

2.2.1.3. Ultrasonography

At present, US is accepted as complementary to clinical examination and mammography with a primary role in distinguishing cystic from solid masses. US is not considered reliable in screening for asymptomatic breast cancer, and its role in differentiating between benign and malignant solid breast masses has been controversial (88-90). Morphologic US criteria for differential diagnosis have, however, been presented recently (90-91). US can be a useful tool when mammography has limitations, as in the identification and evaluation of palpable or nonpalpable masses not visible in mammographically dense breasts (89,90,92). It is also helpful in the case of extremely peripherally located masses, where adequate positioning for mammography may be impossible, as well as in the evaluation of axillary node status and in examining breasts recently operated or acutely inflamed (60). During pregnancy, US is the primary imaging method because mammography has an unacceptably high false-negative rate, probably due to the high radiographic density, and ionizing radiation may be harmful for the developing fetus (93). US also plays an important role in the guidance of diagnostic interventional procedures, including core- and fine-needle aspiration biopsies (94-97).

2.2.1.4. Other methods

The role of new noninvasive diagnostic methods, such as MR imaging or radioisotopic scanning of the breast, has not yet been established. They can be used as a complementary diagnostic modality in preselected patients (98-103). The eventual future indications for MR imaging in breast cancer include a need for more accurate evaluation of size and multifocality (100), postoperative differential diagnosis of tumor from therapy-induced changes (101), and more accurate detection of occult cancer in patients with isolated ipsi- lateral axillary metastasis (102). 18 2.2.2. Invasive methods

2.2.2.1. Needle biopsy

Further investigation of mammographically found suspicious nonpalpable lesions has usually involved surgical WGB with its obvious disadvantages. The lower cost, lesser morbidity and better cosmetic outcome of percutaneous needle biopsy have challenged the superiority of open biopsy in obtaining reliable tissue diagnoses of nonpalpable breast lesions (26,104). As a result, the proportion of benign findings in open surgical biopsies can be markedly reduced (47,105-110), preferably to or below the benign-malignant operation rate of 1:1, when operations for nipple discharge and abscesses are excluded (111). Surgical excisions of nonpalpable breast lesions with preoperatively confirmed malignancy have been associated with larger volume, a high incidence of negative microscopic margins, lower costs and a decreased need for re-excision (27,112,113). According to the European Breast Cancer Working Group, the majority of nonpalpable breast cancers should receive a preoperative diagnosis by FNAB or core needle histology (111). This emphasizes the importance of a preoperative diagnosis allowing thorough pretreatment planning and discussion with the patient. It is important to minimize false-negative needle biopsy findings (114), which may result in a delay of diagnosis and a more advanced stage of breast cancer. In qualified hands, the FNAB technique works reasonably well. In nonpalpable breast lesions, the reported sensitivities range from 57% to 97% and specificities from 91% to 100% (114- 119). For comparison, the corresponding sensitivities and specificities of FNAB in palpable lesions have been 65% to 99% and 64% to 100%, respectively (114). However, FNAB has several limitations. Up to 54% of the cytologic samples are insufficient (106,114,120,121). Recently, the high rate (34%) of insufficient samples for FNAB of nonpalpable breast lesions led to the early termination of one multicenter trial (121). The high risk of insufficient specimens has been related to inexperience with the technique and to the occurence of benign lesions, especially microcalcifications, where it may be difficult to obtain a definitive benign diagnosis (114,116,121). Dedicated breast cytopathologist is needed for all FNABs, and it is still not possible to differentiate between DCIS and invasive ductal carcinoma with a cytologic examination because of the lack of architectural information (114). Histology is preferable to cytology for a definite diagnosis (20). Core-cutting needle biopsy techniques provide sufficient tissue for a histologic diagnosis with a lower rate of insufficient samples (110,122) and have partially replaced FNAB (106,122-125). In nonpalpable breast lesions, reported sensitivities range from 60% to 98%, specificities from 96% to 100% and rates of insufficient samples from 0% to 17% (104,106,119,122,126). Large-core biopsy is more invasive and traumatic than FNAB (127). Yet, the reported complication rate is very low (27,104,126,128). The diagnostic accuracy of image-guided core biopsy varies with the histologic diagnosis, being lower in atypical hyperplasia and DCIS, but notably high in invasive cancer. Atypia may upgrade to DCIS and DCIS to invasive cancer after excision, but DCIS does not downgrade to a benign diagnosis. (104,122,126,129). Therefore, a core-needle biopsy diagnosis of atypia or DCIS mandates an open breast biopsy (104). 19

During the 1990s, technical improvements have been introduced to ensure histologic diagnosis of nonpalpable breast lesions with an accuracy comparable to open surgery. Directional, vacuum-assisted biopsy (Mammotome) uses a vacuum to withdraw and tran- sport larger amounts of tissue more quickly and completely without a need to withdraw the needle each time the biopsy is harvested (128,130). The need to avoid the abovemen- tioned underestimation of disease has been pointed out (131). Stereotactic guidance of fine-needle (20-25 gauge) aspiration cytology or large-core needle (14-18 gauge) aspiration biopsy in nonpalpable suspicious breast lesions is most widely used (105,106,116-119,122,132). The competitive method, US guidance, is less widely documented (95,123,133,134). The different guidance methods have their specific advantages and disadvantages, and they should be regarded as mutually complementary (135). The lesions best suited for stereotactic guidance are suspicious asymmetric densities, sonographically invisible microcalcifications and very small masses in large, fatty replaced breasts. US provides a fast and accurate guidance method without ionizing radiation for virtually all palpable and nonpalpable sonographically visible breast lesions (130). With US, accurate monitoring of the needle tip can be ensured during needle insertion and sampling under continuous real-time visualization (136), which may explain the reported low rate of insufficient specimens by US guidance (121,134,137). US guidance has also been reported to be more cost-effective than stereotactic guidance (138). On the other hand, US-guided biopsies have been claimed to be technically more difficult to perform and, therefore, to have a longer learning curve than stereotactic ones (127).

2.2.2.2. Galactography

Discharge from the nipple is a common reason for remitting a woman to a surgical evaluation, accounting for 3-5 % of all causes of remittance (28-30). Only spontaneous, non- lactational, persistent and unilateral discharge from a single duct is considered to be of surgical significance. Drug-induced or endocrine discharges are usually produced bilate- rally by multiple ducts. Spontaneous discharge from a single duct is a sign of an underlying inflammatory, benign proliferative or even malignant disease. Papillary lesion is the most common etiology in these cases (29,30,139-142). Some controversy exists over the precancerous nature of single intraductal papillomas, but the prevailing view is that multiple papillomas (papillomatosis) may become malignant (32,139). It is also well recognized that pathologic differentiation between papilloma and papillary carcinoma may be extremely difficult. The incidence of cancer in patients in whom nipple discharge is associated with a palpable lump has been up to 60-70 % (30). Without a palpable lump, the reported risk for cancer ranges from 4% to 13% (28-30,81,140-142). On the other hand, only 1-2 % of all breast cancers present solely as nipple disharge (28,81). The sensitivity of mammography is low in these cases (31-33,143). The value of cytologic examination of the discharge is limited, because of the unreliability of exfoliative cytology in the detection of intraductal malignancy (142,144). The best method available for the detection of intraductal tumors, variations in ductal calibre and ductal incontinuities is galactography (ductography) (29,34,81,142,145). An intraductal tumor presents at galactography as a filling defect within a contrast-filled duct (Figure 1) (142). The findings are, 20 however, nonspecific. Differential diagnosis between benign and malignant tumors is not possible. Therefore, it is important to excise all galactographically suspicious lesions for a pathologic examination, even though a pathologic correlate cannot always be found (146).

)LJ$JDODFWRJUDPVKRZLQJILOOLQJGHIHFWVVXVSLFLRXVRILQWUDGXFWDOSDSLOORPDV

0HWK\OHQHEOXHVWDLQLQJ

The mainly benign nature of nipple discharge underscores the importance of precise minimal-volume excision of the intraductal lesions for better cosmesis. Microdochectomy is the conventional method for patients with persistent discharge from a single lactiferous duct (32). Knowledge of the exact location, size and distribution of the duct producing discharge facilitates accurate minimal-volume microdochectomy through a circumareolar incision and ensures inclusion of the lesion in the biopsy specimen (Figure 2). The smaller size of the specimen gives a better cosmetic outcome and provides the pathologist with an intact specimen and fewer sections to review. Methylene blue staining is the met- 21 hod of choice to meet these requirements (142). Other localization methods, such as clinical findings only (30), a lacrimal probe with (29) or without (28,81,147) galactography or with the insertion of a catheter into the secreting duct (148), have a number limitations. Approximately 20% of lesions are missed in duct excision if galactography is the only localization method (34). In 1969, Funderburk and Syphax (139) reported a contrast mammography technique with dye identification of the lobule producing discharge for surgical excision. It has been later criticized (149) for the dye reactions and extravasa- tions seen in a small percentage of patients and for the need for time-consuming coordi- nation between the radiologist and the surgeon. Methylene blue staining lacks these limitations.

Fig. 2. A surgical specimen of a ductal-lobular unit after preoperative methylene blue staining and opening of the duct revealing two intraductal papillomas.

2.2.2.4. Wire-guided biopsy

Today, breast biopsy is one of the most common surgical operations. Despite the advanced image-guided core biopsy techniques, open surgical WGB is still often necessary for a definitive histologic diagnosis of nonpalpable breast lesions. The indications for open biopsy include cases where the mammographic finding is not concordant with the core biopsy results, the mammogram changes or the core biopsy demonstrates atypical hyperplasia (over 50% risk of simultaneous malignancy) (124). The malignancy yield of 22

WGB for nonpalpable breast lesions has been 10-48 % in studies performed outside mammography screening programs (37-42,150-154). An average of 37% of the malignancies have been noninvasive (154). Wire insertion is performed under mammographic or sonographic guidance. The advantages of WGB are adequacy of the specimen and accuracy (155). Specimen radiography is an integral part of WGB to ensure and docu- ment that the mammographic lesion has been partially or totally removed. The failure rates of WGB vary from 0 to 10 % (153,156-158). The rate of complete radiologic excision of all the lesions removed by WGB vary from 47% to 92% (153,159,160). To increase the success rate of WGB, it has been suggested that the radiologist should recommend after specimen radiography that the surgeon should remove more surrounding breast tissue in suspicious cases (60). However, single-view and even two-view specimen radiography have limitations in the accurate verification of positive or negative mammographic margins of a three-dimensional specimen. Apparently clear margins may be found to be involved if an appropriate tangential view is obtained. (159) This problem is further complicated by possible additional pieces of breast tissue removed after specimen radiography. There are no reports concerning the actual contribution of these additional pieces of specimen to the radiologic completeness of WGB. In WGB, because of nonpalpability, the localization wire and the preoperative mammogram constitute the sole guide of tissue excision. Specimen volume in wire-guided breast biopsy is determined by two opposite limitations. The surgeon aims at adequate local control with histologically tumor-free margins without compromising cosmesis with an unnecessarily wide excision of surrounding normal breast tissue. Therefore, the tip of the guide wire should be within 10 mm from the lesion in any plane in at least 80% of cases (111). The reported success rate of the placement of the localization wire within 5 mm of the target lesion has been more than 96% (157). This facilitates crucially the indi- vidualization of the amount of breast tissue excised. More than 95% of nonpalpable mammographic abnormalities should be successfully removed with the first WGB, and 80% of benign breast lesions should weigh less than 30g (111). In the era of active use of image-guided needle biopsy and BCS, a considerable proportion of WGBs also prove to be therapeutic after a histopathologic examination. In cen- ters where this approach has been employed as a definitive therapeutic procedure, 41-60 % of patients require no further local surgery, which results in lower costs and morbidity (156,157,161). The therapeutic role of wire-guided biopsy is, however, still controversial (161,162). 23 2.3. Treatment

2.3.1. Breast-conserving surgery

2.3.1.1. Lumpectomy

In BCS, the casual usage of mutually interchangeable terms, such as "lumpectomy", "partial mastectomy", "wide excision" and "segmentectomy", without specifying the surgical technique in detail has been confusing. However, the term "lumpectomy", as used in this study, has been defined as the complete removal of a tumor with histologically negative margins without regard to any specific anatomic landmarks within the breast, aiming at local-regional control of breast cancer with the resultant appearance of a normal breast (163). BCT for appropriately selected patients with early (stages I and II) breast cancers 4 cm or less in size is equivalent and even preferable to mastectomy, provided that the margins of the resected specimen are free of tumor. Overall survival and distant disease-free survival have been equal in several randomized trials. Only the long-term local failure rates have been different, reflecting the uncertainty with regard to the optimal extent of surgical excision between a high initial local control rate and cosmesis. (21,43-46,164) The use of BCT has increased over time, but there are still great differences between surgeons in the use of this data. The rates of eligible patients receiving BCT have varied from 16 to 80% (165-170). The factors influencing this rate are patient's age, tumor size, urban residence, patient income and surgeon's age and education (165,171-174). There have also been sug- gestions to extend the indications for BCT to certain preselected patients with locally advanced breast cancer (175,176). The use of induction chemotherapy to downstage tumor has resulted in reductions in the size of the primary tumor, allowing BCS in 27- 81% of locally advanced noninflammatory breast cancers with 5-year survival of 73-86% after multimodality treatment (176). Following active preoperative use of image-guided needle biopsy, all biopsies must be performed with an assumption of malignancy, and the procedure of choice should be lumpectomy (177). Therefore, the biopsy should include at least a surrounding 0.5-1.0 cm margin of apparently normal breast tissue (21,50,178,179). With careful preoperative planning and a proper biopsy technique in diagnostic lumpectomy, the rate of re-excision for positive margins has been reported to be as low as 5% (180). The type of anesthesia must be individualized to suit the patient, but either local anesthesia alone, local anesthesia with intravenous sedation or general anesthesia may be used. For wire-guided procedures, however, to achieve free histologic margins in lesions loca- ting deeper than 3 cm in the breast, general anesthesia has been recommended (181). The incision should be small, curvilinear or parallel to the edge of the areola consistent with removal of the lesion following the normal lines of tension in skin (Langer's lines) and least likely to create prominent scars but without compromising the possible mastectomy incision (163,182,183). Transverse or radial incisions can be used in the medial aspect of the breast (163) or directly medial or lateral to the nipple (177). They have, however, been criticized for an inferior cosmetic outcome (178,184), and wide use of circumareolar (periareolar) incisions for breast biopsies and even for total subcutaneous 24 mastectomies and reconstructions has been recommended because of the minimal scar- ring and the small likelihood of keloid formation and deformity involved (184). Howe- ver, some only use this incision directly over the lesion, because tunneling may make dissection difficult and prone to errors (163,182,185). In wire-guided procedures, the placement of the incision over the area of pathology rather than at the skin entry point of the wire allows a small incison with good visualization (186). Skin or subcutaneus tissue are usually not removed with the lumpectomy specimen (3,21). If skin must be removed (as in superficial tumors and in suspicions of inflammatory carcinoma), a radial incision is preferable in the lower hemisphere of the breast (182). After meticulous hemostasis (21), the wound is closed with continuous subcuticu- lar sutures (185) without reapproximation of the underlying breast tissue and without breast drains (21,177), and the cavity is allowed to fill with fibrin and become organized (183). Malignant cells have been claimed to disseminate along hematoma planes and limit the probability of local control (184). Placement of clips or sutures and staining of the periphery of the lumpectomy specimen with India ink aids radiologic and pathologic orientation and evaluation of excision margins. Axillary dissection should always be performed through a separate incision (21,178), even when the lumpectomy incision may only be 3 to 4 cm apart (182).

2.3.1.2. Axillary dissection

In BCT, level I-II axillary dissection is an integral part of the treatment of invasive breast cancer both for staging and for the prevention of axillary recurrence (21). Nodal status guides adjuvant therapy and aids in determining prognosis. However, surgical treatment of the axilla is controversial, and exploration of the axilla can be safely omitted in some patient groups (187,188). In the future, sentinel node biopsy may replace full axillary lymph node dissection in a large number of patients (189,190).

2.3.1.3. Margins, residual disease and local recurrence

Numerous reports have been published on the importance of adequate tumor-free margins in lumpectomy specimens (191-207). Most of these reports have focused on the correlation between surgical margins and local recurrence (195-198,200-205,208,209) and some on the occurrence of post-biopsy residual cancer (192-194,210). In BCT, margin involvement has been associated with tumor size (206,211,212), EIC (207,211,212), age (211), mammographic findings (212) and the type of operation (212). Positive or unknown histologic margins have correlated with the finding of residual disease at re-excision in 29-65% of cases (180,193,194,213,214) and less commonly also with a variety of other factors, such as EIC, residual microcalcifications on post-biopsy mammograms, clinical tumor size, method of detection, pathologic nodal status and histologic appearance of the tumor (192,194,210,213,215). When the first attempt to provide histologically tumor-free margins fails, re-excision may be necessary. Residual microscopic carcinoma after BCT accounts for the majority of recurrences occurring early and close to 25 the site of previous excision in the breast (191-193,216-218). Thus, there is a growing desire to ensure the removal of all tumor tissue at lumpectomy. The reported proportions of residual carcinoma in re-excision specimens have varied according to the relative extent of the primary biopsy procedure from 11 to 100 % (191-195,210,219-221). They have also been related to the extent of secondary surgery (breast-conserving surgery vs mastectomy) (206). Currently, there is no firm consensus on the optimal extent of the initial surgical procedure in BCT. The data concerning the need to obtain microscopically negative margins are conflicting. Several reports have shown a correlation between positive margins and local recurrence (195-198,202-205), but it has not been confirmed consistently (200,201,208,209). However, even Solin et al. (200) from the latter group recommend re- excision to achieve microscopically negative margins before irradiation therapy. The general recommendations for the histologic width of the microscopic resection marginals vary from "free margins" up to 2 cm (3,21,111,169,179,222,223). However, unnecessarily wide margins (> 2 cm) should be avoided for cosmetic reasons (21). The European Breast Cancer Working Group requires only clear margins in the surgical management of nonpalpable breast cancer (111). Positive histologic margins mandate re-excision, even though re-excision predicts an adverse cosmetic outcome (224). In addition to positive excision margins of the primary tumor, the predictors of risk for local recurrence include lymphatic vessel invasion, low tumor grade and large tumor size (218). EIC has also been considered prognostic for local failure after BCT, but the results are controversial (191,195,196,199,201,203-205,216). The detection rate of local recurrence following BCT increases with the length of follow-up. In an overview of the randomized trials, the addition of radiotherapy to surgery has reduced the rate of local recurrence by two thirds (225). Fisher et al. (44) reported, after 12 years of follow-up, a local recurrence rate of 35% following BCS alone and 10% following BCT. Notably, all these series include palpable invasive breast cancers, whereas corresponding data concerning nonpalpable and noninvasive breast tumors are very sparse. According to Silverstein et al. (202) patients with ductal carcinoma in situ in whom the margin width is less than 1 mm can benefit from postoperative radiation therapy. Excellent local control can be achieved without radiation therapy when excision margin width of at least 10 mm are obtained. Earlier reports have also shown the difficulties encountered in assessing margins by a standard histologic examination, which is probably the main reason why margin analysis is a poor predictor of the completeness of tumor excision (44,193,211,213). Frazier et al. (193) pointed out that the assessment of margins may be misleading, since 48 % of patients with involved margins had no residual tumor, and 30 % of those with clear or close margins subsequently had residual tumor upon re-excision. Breast cancer usually involves a more or less round tumor that subclinically spreads radially and imperceptibly into the surrounding breast parenchyma, making the determi- nation of histologic margins very difficult. At least two thirds of invasive breast cancers have root-like borders, and this proportion is increased by the presence of associated intraductal carcinoma. (87,217) This greatly and unpredictably increases the tumor surface area to be histologically examined. The most common reason for a false-negative pathologic report of lumpectomy margins is inadequate examination, because only a very small proportion of the specimen's outer surface can be analyzed. It has been calculated that a complete display of the surface of a 2-cm spherical carcinoma would require more than 26

3,000 histologic sections, which is not feasible. (217) Thus, success is obviously related to the number of sections assessed for margin involvement. Also the discrepancy between a large tumor size and the standard pathological slides measuring 2.5 x 3 cm at maximum may cause problems in a histopathological examination. Moreover, pathologists' examinations of specimens are far from uniform, and there is considerable variation in the definition of histologically positive margins in different reports. In the National Surgical Adjuvant Breast Project (NSABP protocol 6), the headquarters' pathologist agreed with the referring pathologists on only one third of the cases with positive histologic margins. According to the NSABP study, it is most appropriate to regard the lines of resection involved only when cancer is transected. (191) Because of these problems, alternative methods for assessing surgical margins have been suggested. Carter (217) proposed peeling of the entire outer surface of the lumpectomy specimen after inking and parallel sectioning of it for histologic evaluation. Fisher et al. (191) performed sectioning of lumpectomy specimens in several different planes to topographically localize tumor cells in representative margins. Margins have also been evaluated by histologic examinations of tumor bed biopsies and a fully excised cavity wall (211), frozen sections of surgical margins (226) and cytologies of samples obtained by scraping (227) or shaving (228) from the cavity walls or from the outer surface of the lumpectomy specimen (229). Theoretically, the cytologic examination evaluates the entire resected surface even intraoperatively (229). Unfortunately, the results of these methods have either been controversial or there has been a lack of confirmation of preliminary results. However, the most promising method has been the touch preparation cytology introduced by Cox et al. (229) in 1991, which is a rapid intraoperative method for evaluating the entire resected breast lumpectomy surface. The authors have recently related their cytologic evaluation of lumpectomy margins with local recurrence 3 to 5 years postoperatively (230). If re-operation is to be avoided, it is mandatory to assess the margins intraoperatively. Frozen sections are widely used for this purpose, but their usefulness has been limited by the relatively high proportion of false negatives and deferred diagnoses (231). Employing cytologic examination in conjunction with frozen sections during surgery has been claimed to increase the validity of margin assessment (87). However, the use of frozen sections has been discouraged for small (< 10 mm) and nonpalpable breast lesions because of the fear that manipulation may crush the lesion undetectable, and permanent sections from the frozen block may be suboptimal or even useless for definitive diagnosis (3,232).Thus, no optimal method for assessing the microscopic margins of a lumpectomy specimen has been established. In WGB for nonpalpable breast cancer, specimen radiography is used for intraoperative assessment of tumor-free margins. The accuracy of the method in predicting histologic margins of the specimen or residual disease in re-excisions has, however, proved limited (159,160). The incidence of involved histologic margins after WGB ranges from 25% to 84% (151,157,159,160,233,234). Campbell et al. (234) aimed at a margin of 1 cm around the lesion and reported the lowest rate of positive margins. Thus, the wide range is probably attributable to the impact of the relative extent of biopsy, as previously noted in series including palpable tumors (207,208). In WGB, the reported rates of residual cancer at re-excision after positive histologic margins vary from 22 % to 64 % (151,160,162,234). 27 2.3.1.4. Multifocality and multicentricity

Assessment of completeness of lumpectomy is further complicated by the multifocality and multicentricity of tumors. In the literature, the terms "multifocality" and "multicentricity" have been confusingly used interchangeably. For clarity, microscopic foci of cancer encountered within or close to the same quadrant as the index cancer should be regarded as signs of multifocality and those seen in quadrants remote from the index cancer as signs of multicentricity. Breast cancer is a multicentric disease. In the reports from the previous era, when mastectomy was a common treatment, the rate of multicentricity in nonpalpable breast cancer varied from 34% to 47% (235-239). According to Holland et al. (240), tumor resection with 3 to 4 cm margins would have left invasive or noninvasive residual cancer in the remaining breast tissue in 4 to 9% of cases. According to Morimoto et al. (241), lumpectomies removing a T1 tumor with a margin of 2 cm would leave residual cancer foci in the breast in about 20% of cases. Both of theses series included palpable tumors. It has been estimated that even in the presence of apparently negative histologic margins, about one third of breast cancers are multifocal or multicentric beyond the limits of local resection (217). Thus, complete excision of some breast cancers with lumpectomy may be difficult or impossible. Local recurrence is usually noted near the primary site, and it is often concordant with the histologic type and grade of the index tumor. This suggest that local recurrence represents the growth of an overlooked multifocal tumor, minimizing the clinical significance of multicentricity. (191,218) Multifocality or multicentricity of breast cancer is, however, not considered an absolute contraindication for performing BCS (191). In BCT, multifocal disease should preferably be removed by lumpectomy, and subclinical multicentric involvement may subsequently be eradicated by irradiation (87). However, postoperative irradiation therapy only prevents approximately two thirds of local recurrences after BCT (225).

2.3.2. Mastectomy

The current indications for mastectomy in breast cancer are extensive malignant type calcifications visible on the mammogram, multicentricity, inflammatory breast cancer, failure to obtain tumor-free margins, contraindications for radiotherapy (e.g. physical disabili- ties, pregnancy, previous irradiation of the breast), large tumor size in proportion to breast size and the patient's preference for mastectomy (3,177). Controversy exists about such factors as EIC, extensive lymphatic involvement and age under 39 years (21). However, mastectomy is still performed in several patients eligible for BCT for a variety of reasons (165,171-174) (see paragraph 2.3.1.1.). The reported rate of mastectomies in the surgical treatment of early breast cancer varies between 20% and 84%, showing a decreasing ten- dency over time (129,165,167-170). 28 2.3.3. Cosmesis

There are numerous reports on the cosmetic outcome after BCT, showing most patients to have good or excellent overall results (224,242-250), but the data concerning the cosmetic outcome after biopsies of benign breast lesions are very sparse. No reports on the cosmetic outcome after WGB of breast lesions that subsequently proved benign could be found. A considerable proportion of nonpalpable lesions in WGB subsequently prove benign, underscoring the importance of the cosmetic result of this widely-used procedure. The fear of poor cosmesis may delay treatment, which can be tragic whenever a malignant tumor is involved. The adverse impact of irradiation after BCS on the overall cosmetic outcome is almost impossible to evaluate. It is notable (224,242-250), however, and such changes are absent after WGB for beningn breast lesions. The surgical changes in the breast tend to be evident within months after lumpectomy (185), because human scar tissue maintains a relatively high level of collagen synthesis for only up to 6 months (251), but the cosmetic scores after BCT continue to deteriorate for the first 36-60 months because of postradiation changes (243,244,252). The development of hematoma, seroma, and subsequent fibrosis at the excisional biopsy site after BCT have been related to lower cosmetic scores in a few reports (224,247). Retraction due to scar tissue is further aggravated by radiation- induced fibrosis (253). Several investigators have reported a relationship between increasing specimen volume and inferior esthetic outcome in BCT (224,242-249). Some of these reports show a strong correlation between any volume resected and a lower probability of excellent cosmetic outcome (243), but some only show such correlation for large volumes of breast tissue (>70, >85, or >200 cm3) (244,248,249,254). Large specimen volumes in BCT are probably related to a large tumor size and the doses of irradiation administered. Other factors, such as age, breast size, location of the lesion and body mass index have been related to lower cosmetic grading in some reports (242,243,245,246,250,255), but the results have been controversial. 3. Aims of the study

The purpose of the present work was to evaluate the results of ultrasonographically guided fine-needle aspiration biopsy (US-FNAB), preoperative methylene blue staining, wire-guided biopsy (WGB) and lumpectomy in suspicious nonpalpable breast lesions and early breast cancer in Oulu University Hospital, aiming to increase the success rate in obtaining radiologically and histologically free margins without residual disease in breast cancer. The specific aims of the study were: 1. To study the usefulness of US-FNAB in nonpalpable suspicious breast lesions in clinical practice. (Paper I) 2. To evaluate the preoperative use of methylene blue staining of nonpalpable galactographically suspicious breast lesions. (Paper II) 3. To investigate the determinants of positive histologic margins and residual cancer in WGB of nonpalpable breast cancer and in lumpectomy for early breast cancer and the determinants of positive radiologic margins and the correlation between radiologic and histologic margins and residual disease in WGB. (Papers III and IV) 4. To study the usefulness of touch preparation cytology in the examination of lumpectomy margins for early breast cancer. (Paper IV) 5. To evaluate the cosmetic outcome and related determinants after WGB for benign proven nonpalpable lesions. (Paper V) 4. Materials and methods

4.1. Patients

The patients discussed in paper I consisted of 86 women with 90 nonpalpable suspicious breast lesions subjected to sampling by US-FNAB at Oulu University Hospital between February 1986 and December 1993. The cases were derived from the files of all patients (n=781) who underwent wire-guided breast biopsy for histologic diagnosis because of a suspicion of malignancy after mammographic and/or sonographic examination. The lesions were detected through screening mammography in 61 patients, evaluation of nonspecific symptoms in 19, and cancer follow-up in six. Paper II considers a series of 30 consecutive women who underwent breast operations at Oulu University Hospital from 1986 through 1995 for persistent unilateral nipple discharge from a single nonlactiferous duct without a palpable mass in the breast (mean age 59 years; range 31-75 years). The discharge was bloody in 22 (73 %) patients and serous in 8 (27 %) patients. The mammography of 28 (93 %) patients was considered normal. A small retromamillary density and retromamillary fibrosis were the mammographic findings in the remaining two patients. Galactography of the suspicious duct was attempted preoperatively in all cases, and methylene blue staining was performed on the day of surgery. The 69 patients (mean age 54 years; range 31-82 years) with 71 nonpalpable breast cancers studied for paper III and the 101 patients (mean age 50 years; range 18-72 years) with subsequently benign proven nonpalpable breast lesions without previous surgical biopsy of either breast covered for paper V were prospectively derived from the series of 232 consecutive wire-guided breast biopsies performed between September 1993 and June 1996 at Oulu University Hospital, in Finland. Paper IV covers 53 consecutive prospectively evaluated patients (mean age 56 years; range 40-82 years) who underwent 55 lumpectomies for early breast cancer at Oulu Univer- sity Hospital from October 1994 until November 1995. Two lumpectomies were re-excisions. Thirty-six of the 55 lumpectomies (65.5 %) were perfomed under wire guidance and radiologic control. 31 4.2. Methods

4.2.1. US-FNAB in nonpalpable breast lesions (I)

US was performed after the primary mammographic examination using the direct contact method and real-time equipment (Toshiba SSA 77 or 100, Aloka SSD-650, or Quantum 2000) with a 5 or 7.5 MHz transducer. After the lesion had been localized sonographically, it was punctured and aspirated under real-time control with the free-hand technique using a Cameco® syringe holder connected to a 10 ml syringe and a 0.6 or 0.7 mm diameter needle. At least two aspirates were taken from each lesion to obtain a sufficient amount of material for cytologic examination. The samples were fixed in 50 % alcohol and stained with Papanicolaou stain. The findings were classified according to the Papa- nicolaou method into classes 0 to V. Class 0 indicates an acellular or insufficient aspirate; class I indicates a normal finding with no atypical or abnormal cells; class II indicates abnormal cells with no evidence of malignancy; class III shows atypical cytologic results in which malignancy cannot be excluded; class IV is probably malignant; and class V shows malignant cytologic results.

4.2.2. Preoperative methylene blue staining (II)

Before galactography, the secreting duct was identified by stroking the breast and squee- zing the nipple. The breast was cleaned with klorhexol 5 mg/ml solution and the duct was cannulated with a thin (0.5-0.8 mm) blunt-ended needle connected to a syringe contai- ning water-soluble contrast medium. 0.1-2 ml of contrast was injected until a mild resis- tance was noted or the patient felt uncomfortable. Magnified mediolateral and craniocaudal mammograms were obtained with a dedicated microfocus tube mammographic unit (Mamex DC Mag or Ami) to visualize the secreting duct. On the day of surgery, galactography was repeated using 1:1 solution of contrast medium and 1 % methylene blue. Mediolateral and craniocaudal mammograms were obtained to ensure that the same duct was stained as in the primary galactography. After successful staining, it was easy to localize and dissect the entire duct system and part of the lobule for a permanent section through a circumareolar incision.

4.2.3. WGB for nonpalpable breast cancer (III)

A radiologist performed the preoperative localization of the lesion using mammographic or sonographic guidance on the morning of surgery and later checked the specimen radiography in all cases to ensure adequate excision. The technique of WGB was compa- tible with the cosmetic principles recommended in the literature, and skin incisions least likely to create prominent scars were used (182). The biopsy incision was made directly 32 over the tip of the guide wire. The aim was to remove the whole lesion with a margin of 0.5-1.0 cm of apparently normal breast tissue. The specimen was oriented with metallic clips. The volume of the surgical specimen was measured by water displacement in a glass measure. The measurement was accurately performed in 68 cases. A magnified two-view radiograph of the specimen was obtained using mild compression. After identifying the lesion on the specimen radiograph, the smallest tumor-free margin was measured. The resection margins of the excised specimen were stained with India ink, dried, fixed in for- malin, and serially sectioned at 2-3 mm for a histologic examination. A microscopic margin was considered to be positive if the tumor was present within the inked surface, and otherwise it was considered negative. Re-excision of the primary tumor bed was performed in 54 (76.1 %) of the 71 cases (mastectomy 28, breast-conserving surgery 26) according to our current policy (histologic margins <5 mm). In some small tumors, however, the biopsy site was not re-excised in spite of only 2-4 mm histologic margins. Moreover, in one small DCIS with involved margins no re-excision was carried out. EIC was considered to be present when in situ cancer occupied 25 % or more of the area encompassed by the infiltrating tumor or was widely present in random sections of grossly unremarkable breast tissue, or clearly exten- ded beyond the infiltrating edge of the tumor into the surrounding breast tissue. Table 1 shows the recorded variables.

Table 1. The recorded variables for determinants of positive radiologic and histologic margins and residual disease in re-excisions in breast cancer patients.

Variables Age Menopausal status Body-mass index Breast size Mode of detection Mammographic finding Mammographic size Method of wire-guidance Number of wires Point of entry into breast Anesthesia Directionofskinincision Length of skin incision Quadrant of breast incision Depth of incision Specimen volume Histology of lesion EIC Pathologic size Multifocality Multicentricity 33 4.2.4. Margins and residual tumor after lumpectomy (IV)

In lumpectomy, the surgeon removed the tumor with a rim of grossly normal tissue, usually 1-2 cm or more. The following parameters of the 53 consecutive patients with 55 lumpectomies were recorded: age, preoperative palpatory and mammographic findings, size of the biopsy specimen (largest dimension accurately measured in 51 of the 55 specimens) and pathology of the index tumor (size, stage, presence vs. absence and extent of noninvasive cancers and multifocality). During surgery, cellular samples were obtained for touch preparation cytology as described by Cox et al. (229). Two to six slides were used, depending on the specimen size. After that, the specimen was inked for a permanent histologic examination.The touch preparation cytology slides were categorized after examination as positive, if malignant cells were present, negative, if malignant cells were not seen, and inconclusive, if the cellular material was inadequate to allow conclusions. A microscopic margin was considered positive if the tumor was present within the inked surface, but otherwise it was considered negative. The distribution of the histologic types of breast cancer is shown in Table 2. Twenty- three of the 55 specimens (41.8 %) contained pure invasive and 10 pure noninvasive (in situ) tumor, while the remaining 22 cases contained both invasive and noninvasive tumor. In seven of these 22 cases the noninvasive component was considered extensive (EIC). The mean (SD) pathologic size of tumors (invasive + noninvasive component) was 1.8 (1.4) cm. According to UICC (256), the pathologic stages 0, I and II accounted for 10 (18.2 %), 30 (54.5 %) and 15 cases (27.3 %), respectively. Seven of the 55 tumors (12.7 %) were multifocal. Re-excision of the primary tumor bed was performed in 34 of the 55 cases (61.8 %) because of uncertainty as to the adequacy of the surgical margins (5 mm or less) in the initial procedure. The same parameters were compared in the cases with and without residual cancer in the re-excision specimen.

Table 2. Distribution of histologic types of breast cancer in 55 lumpectomy specimens.

Histology N Invasive Ductal 31 (12) Lobular 5(4) Tubular 5(5) Mucoid 2(2) Papillary 1 Metaplastic 1 Noninvasive Ductal 6(4) Lobular 4(4)

Total 55 (30)

The figures in parentheses indicate the number of nonpalpable tumors. 34 4.2.5. Cosmesis (V)

The recorded potential risk factors for poor esthetic outcome are shown in Table 3. The cosmetic outcomes were evaluated 6 months postoperatively by using a subjective (observer-based) method. All ratings were based on a comparison of the treated breast with the untreated breast (with the exception of the surgical scar) by using a modification of the European Organisation for Research and Treatment of Cancer (EORTC) breast cancer cooperative group protocol 10801 rating of cosmetic results of breast-conserving treatment (257). Loss of breast tissue, breast contour from the side, breast contour from the front and shift of the areola and nipple were scored as follows: 0 = no difference, 1 = a small difference, 2 = a moderate difference, 3 = a large difference. Scar retraction and the surgical scar in itself were scored as follows: 0 = unobtrusive, 1 = visible but not affecting the cosmetic results, 2 = visible and detracting somewhat from the cosmetic results, 3 = visible and detracting notably from the cosmetic results. The sum of the ratings could vary from 0 to 18. They were classified into four descriptive subdivisions: excellent 0-1; good 2-3; fair 4-5; and poor 6 or more. A single observer (AOS) evaluated 94 patients and two other specialists the remaining 7 patients. All patient appraisals were recorded using the same 4-grade scale.

Table 3. Recorded risk factors for inferior cosmesis.

Risk of factors Menopausal status Body mass index Mammographic finding Radiologic size of lesion Method of wire guidance Number of wires Point of entry into breast Anesthesia Amount of local anesthetic Directionofskinincision Length of skin incision Quadrant of breast incision Depth of incision Subcutis excised with specimen Skin excised with specimen Intraoperative blood loss Specimen volume Closure of wound Complications Histology of lesion 35 4.2.6. Statistical methods

Multivariate logistic regression analysis was chosen to identify the risk factors associated with positive mammographic and histologic margins and residual disease in re-excision specimens (Paper III) or with unsatisfactory cosmetic results (Paper V) . For clarity, the dependent variable in Paper V was re-categorized into two classes: satisfactory (excellent / good) and unsatisfactory (fair / poor). All the categorical independent variables were cross-tabulated against the dichotomous results and the associations evaluated using the chi-square test. The continuous variables were analyzed using unpaired t-test. The variables possibly (p ≤ 0.1 in Paper III; p < 0.2 in Paper V) associated with positive margins or residual disease (Paper III) or with unsatisfactory cosmetic result (Paper V) in this univariate analysis were then included in a multivariate logistic regression analysis. A p- value < 0.05 was required for a variable to be included in the final model. Multivariate odds ratios and the respective 95% confidence intervals (95% CI) are shown. McNemar's test was used for a comparison of the cosmetic outcome with the patient's appraisal in Paper V. Chi-square analysis with Yates correction was used for the comparison of categorical data, and Student's t-test was applied for continuous variable comparisons in Paper IV. Differences at or below the level of p = 0.05 were regarded as statistically significant. The calculations were perfomed with SPSS 7.5 software (SPSS inc., Chicago, Illinois, USA) for personal computers. 5. Results

5.1. US-FNAB in nonpalpable brest lesions (I)

The histological diagnoses of the 90 nonpalpable breast lesions are shown in Table 4. The proportion of malignant lesions was 29%, with three intraductal cancers included. One malignant breast lesion proved to be a metastasis from a malignant melanoma in the foot.

Table 4. Histological diagnosis of 90 nonpalpable breast lesions.

Histology No. of lesions Malignant Invasive ductal cancer 22 Intraductal cancer 3 Metastasis of melanoma 1 Benign Fibroadenoma 26 Fibrocystic disease 19 Atypical hyperplasia 9 Fibroadenosis 4 Intraductal papillomatosis 3 Fibrosis 2 Ductal hyperplasia 1

The mean sonographic diameter was 12 mm (range 4-50 mm) in the 68 lesions for which an accurate diameter was available. A comparison of the cytologic and histologic results is shown in Table 5. The cytologic classes I and II were combined as benign, and class III was regarded as neither positive nor negative in the calculations. The sensitivity of this technique in the diagnosis of cancer was 84%, specificity 93%, positive predictive value 94%, negative predictive value 95% and overall diagnostic accuracy 90%. All of the insufficient specimens were included in these calculations. 37

Table 5. Comparison of cytologic results of US-FNAB and histologic results of WGB.

Histologic Results Cytologic results Atypicalhyper- Invasive Benign DCIS* Total plasia cancer Insufficient1101 3 Benign 33 6 0 2 41 Atypical 20 2 2 5 29 Probably malignant 1 0 1 6 8 Malignant0009 9

Total 55 9 3 23 90

*Intraductal cancer

Two false-negative results and one insufficient sample were found among the 26 malignancies. The two cancers with a false-negative cytologic result had sonographic dia- meters of 5 and 8 mm, while the mean sonographic diameter for all cancers (with 22 of the 26 cases accurately measured) was 10 mm (range 5-17 mm) . The ultrasonographic size of the lesion with an insufficient cytologic sample was 10 mm. When these two false- negative cytologic samples were reviewed by a cytopathologist, one (class I) was rede- fined as clearly insufficient, but the diagnosis of the other (class II) remained unchanged. The 64 benign lesions included one false-positive and two insufficient samples. The false-positive case (class IV) was a fibroadenoma, and it was reconsidered atypical or indeterminate in a cytopathologic review. The histologic results of the atypical cytologic samples (class III) are shown in Table 6.

Table 6. Histology of the 29 atypical cytologic aspirates.

Histology Number of lesions Malignant Invasive ductal cancer 4 Intraductal cancer 2 Metastasis of melanoma 1 Benign Fibroadenoma 8 Fibrocystic disease 7 Atypical hyperplasia 2 Fibroadenosis 2 Intraductal papilloma 1 Fibrosis 1 Lactatic proliferation 1 38 5.2. Preoperative methylene blue staining (II)

Galactography was successful in 29 out of 30 cases (96.7%), showing an intraductal tumour in 23 cases (one with discontinuity), ductectasia in two, discontinuity in two and ductectasia with discontinuity in the remaining two cases. Table 7 summarizes the histologic results in the different galactographic findings. Preoperative methylene blue staining was successful in 22 out of 30 (73.3%) cases, and these cases were operated on by standard microdochectomy. In five cases the stained duct was different from the duct seen at primary galactography, and in three cases the cannulation of the duct itself failed. The success rate of methylene blue staining in the different galactographic findings is presented in Table 8. All the eight cases with unsuccessful methylene blue staining were treated operatively, four by quadrantectomy and four by a smaller resection based on clinical, previous galactographic and peroperative findings.

Table 7. Histologic results of 30 patients with different galactographic findings.

Galactography Histology N (%) Intraductal papillomatosis 10 (33.3) Intraductal papilloma 10 (33.3) Intraductal tumor Invasive ductal cancer 1 (3.3) Lobular cancer in situ 1 (3.3) Fibrocystic disease 1 (3.3) Intraductal papillomatosis 1 (3.3) Ductectasia Fibrocystic disease 1 (3.3) Intraductal papillomatosis 1 (3.3) Ductectasia with discontinuity Intraductal papillom 1 (3.3) Discontinuity Fibrocystic disease 2 (6.6) Failure Intraductal papillomatosis 1 (3.3)

Table 8. Success of methylene blue staining in 30 patients with different galactographic findings.

Methylene blue Galactography Success Failure Intraductal tumor 18 5 Ductectasia 1 1 Ductectasia with discontinuity 2 0 Discontinuity 1 1 Failure 0 1

Total 22 8 39 5.3. WGB for nonpalpable breast cancer (III)

Twenty-seven invasive ductal carcinomas, 19 ductal carcinomas in situ, 15 invasive lobular carcinomas, three lobular carcinomas in situ, six tubuloductal carcinomas and one mucinous carcinoma were the dominant histologic subgroups of the lesions revealed by the microscopic examination. The mean pathologic size (invasive + noninvasive component) in the biopsy specimen was 14 (range 1-80) mm. Multifocality was seen in 19 (27%) of the 71 cases. The initial specimen radiography showed the target lesion in 68 (96 %) of the 71 cases, and the remaining three lesions were seen in the synchronously excised additional pieces of specimen. Moreover, one or two additional pieces of specimen were taken in 14 other biopsies after specimen radiography aiming at complete removal of the lesion. The radiologic margins of the excised piece of tissue involved tumor in 19 (27 %) of the 71 cases according to the specimen radiography. The mean radiographic margin was 7 (0-20) mm. In univariate analysis, large mammographic (p=0.02) and pathologic (p=0.04) size were risk factors for positive radiologic margins, but only large mammographic size proved to be significant in multivariate analysis (p<0.01) (odds ratio 1.07; 95 % confidence interval 1.02, 1.12) The mean mammographic sizes of the margin-positive and negative lesions were 28 (7-100) mm and 14 (4-40) mm, respectively. The mean histologic margin was 2 (0-10) mm. In twenty-nine (41 %) of the 71 cases the histologic margins involved tumor. Five variables (p ≤ 0.1 in univariate analysis) were forced into the multivariate analysis, in which superficial excision, large pathologic size and multifocality were significantly related to positive histologic margins (Tables 9 and 10). Residual cancer at or near the biopsy site was found in 13 (24 %) of the 54 cases with re-excision (invasive 4, noninvasive 9). Multicentricity was found in five cases. Five variables (p<0.1 in univariate analysis) were forced into the multivariate analysis, where only multifocality was significantly associated with residual cancer in re-excisions (Tab- les 11 and 12).

Table 9. Risk factors for positive histologic margins in 71 wire-guided biopsies for nonpalpable breast cancer.

Univariate Multivariate Multivariate odds Risk factors analysis model*** ratio ( 95 % CI) Mammographic finding* 0.008 Mammographic size** 0.05 Depth of excision* 0.03 0.045 0.3 (0.1-1.0) Pathologic size* 0.005 0.007 1.1 (1.0-1.2) Multifocality* 0.00001 0.0003 19.8 (3.9-100.0)

*Chi-square analysis. ** Unpaired t-test, non-equal variances. ***Goodness-of-fit test p=0.52. 40

Table 10. Distribution of depth of excision, pathologic size and multifocality according to histologic margins in 71 wire-guided biopsies for nonpalpable breast cancer.

Histologic margin Risk factor Positive Negative Depth of excision Superficial 11 15 11 Down to fascia 14 31 14 31 Pathologic size (invasive +noninvasive) (range) 21 (3-80) mm 10 (1-40) mm Multifocality No 13 39 Yes 16 3

Table 11. Risk factors for residual cancer in 54 re-excisions after 71 wire-guided biopsies for nonpalpable breast cancer.

Multivariate Multivariate odds ratio Risk factor Univariate analysis model*** (95 % CI) Menopausal status* 0.09 Mode of detection* 0.039 EIC* 0.06 Pathologic size (invasive+noninvasive)** 0.085 Multifocality* 0.00001 0.0001 32.1 (5.6-182.4) *Chi-square analysis. ** Unpaired t-test, non-equal variances. ***Goodness-of-fit test p=0.50

Table 12. Distribution of multifocality according to residual disease in 54 re-excisions in 71 wire-guided biopsies for nonpalpable breast cancer.

Residual cancer Multifocality Present Not present Yes 11 6 No 2 35

The relations between margins in specimen radiography and histologic examination and residual disease in re-excisions are shown in Tables 13 and 14. 41

Table 13. Corelations between specimen radiographic and histologic margins and residual disease in re-excisions. The figures in parentheses indicate multifocal tumors.

Radiologic vs histologic Radiologic margins vs Histologic margins vs Correlations margins residual disease residual disease True-positive 11 3 11 (9) False-negative 18 8 2 (2) False-positive 8 6 17 (6) True-negative 34 22 24 (2)

Total 71 39** 54 ( *Additional pieces of specimen (n=17) excluded. **Specimens with additional pieces (n=17) and without re- excision of biopsy site (n=15) excluded.

Table 14. The accuracy values between specimen radiographic and histologic margins and residual disease in re-excisions.

Radiologic vs histologic Radiologic margins vs Histologic margins vs Accuracy margins residual disease residual disease (%) (%) (%) Sensitivity 38 27 85 Specificity 18 79 59 Positive predictive value 58 33 39 Negative predictive value 65 73 92 Overall diagnostic accuracy 63 64 65

5.4. Margins and residual tumor after lumpectomy (IV)

Histologic examination of the specimens revealed positive margins in eight (14.5 %) and negative margins in 47 (85.5 %) of the 55 cases. The correlation between histologic and touch preparation cytologic resection margins is shown summarized in Table 15. The sensitivity, specificity, positive and negative predictive value and overall diagnostic accuracy of touch preparation cytology were 37.5 %, 85.1 %, 37.5 %, 88.9 % and 78.2 %, respectively. The two inconclusive specimens were included in these calculations. The presence of DCIS (with or without an invasive component) in the index tumor predicted positive histologic margins (p=0.03) (Table 16). The mean (SD) pathologic sizes (invasive + noninvasive component) of the margin-positive and margin-negative tumors were 2.8 (1.7) cm and 1.7 (1.3) cm, respectively (p=0.02). 42

Table 15. Relations between histologic and touch preparation cytologic resection margins in 55 specimens.

Histologic margins Cytologic margins Positive Negative Total Positive 3 5 8 Negative 5 40 45 Inconclusive 0 2 2

Total 8 47 55

Table 16. Presence of ductal carcinoma in situ (with or without an invasive component) in eight tumors with positive and 47 tumors with negative histologic margins after lumpectomy.

Presense of DCIS* Margin positive Margin negative Yes 7 18 No 1 29

Total 8 47 * Ductal carcinoma in situ

A histologic examination of the 34 re-excision specimens (BCS 18, mastectomy 16) revealed residual disease at or near the primary resection margin in seven cases (20.6 %) and no residual carcinoma in 27 cases (79.4 %). All the seven re-excision specimens with residual disease contained noninvasive cancer, while two contained invasive cancer. There was no significant difference in the histologic or cytologic margin status between the cases with or without residual cancer (p=0.39 and p=0.80, respectively) (Table 17). The re-excision specimens contained residual carcinoma in 37.5 % and 33.3 % of the cases with histologically and cytologically positive margins, respectively. For negative margins, the corresponding figures were 15.4 % and 18.5 %. Considering all the cases in which either histologic or cytologic margins were positive, only 32.4 % of the re-excisions revealed residual disease. 43

Table 17. Comparison of histologic and cytologic margins to residual cancer found in re- excision specimens.

Residual cancer Total Histologic margins Yes No Positive 3 5 Negative 4 22 26 Total 7 27 34 Cytologic margins Positive 2 4 6 Negative 5 22 27 Inconclusive 0 1 1 Total 7 27 34

Table 18 summarizes the significant differences in the seven index tumors that left residual cancer in the tumor-bed compared to the 27 tumors without residual disease. The differences in the distribution of other factors were statistically insignificant.

Table 18. Main differences between seven tumors with and 27 tumors without residual cancer in re-excision specimens. The figures in parentheses indicate nonpalpable tumors.

Residual cancer Findings in re-excised cases Yes No Multifocal Yes 4(4) 1(1) p=0.003 No 3(3) 26 (12) Palpable Yes 0 14 p=0.040 No 7 13 Preoperative mammography microcalcifications 5(5) 6(6) mass or density 1(1) 16 (6) p=0.07 MC*+density 1(1) 3 normal 0 2 (1)# *Microcalcifications. #Detected at ultrasonography.

5.5. Cosmesis (V)

Overall, the cosmetic outcome was rated as excellent in 49 (48.5%) patients, good in 27 (26.7%), fair in 13 (12.9%) and poor in 12 (11.9%). After re-categorization of the cosmetic outcome into satisfactory (excellent / good) and unsatisfactory (fair / poor), five variables (p<0.2 in unvariate analysis) were forced into the multivariate analysis, in which excision down to pectoralis fascia and postoperative complications showed signifi- 44 cant adverse impacts on cosmesis. According to the goodness-of-fit test, the predicted values fitted quite well to the original data (Table 19). The distribution of these risk factors according to the cosmetic result is shown in Table 20.

Table 19. Risk factors for unsatisfactory (fair / poor) cosmetic results after 101 wire- guided biopsies of benign breast lesions.

Univariate analysis Multivariate Multivariate odds Risk factors (p-value) model*** ratio (95 % CI) Anesthesia (general / local)* 0.07 Depth of excision (superficial / deep)* 0.0003 0.001 6.8 (2.2 to 20.9) Intraoperative blood loss (=<50 / >50 ml)* 0.14

Specimen volume (cm3)** 0.1 Complications* 0.013 0.018 3.8 (1.3 to 11.3) *Chi-square analysis. **Unpaired t-test, non-equal variances. ***Goodness-of-fit test p=0.38 of the final model

Table 20. Distribution of depth of excision and complications according to cosmetic outcome after 101 wire-guided biopsies of benign breast lesions.

Satisfactory (excellent / Unsatisfactory (fair / Univariate odds ratio good) poor) (95 % CI) (n=76) (n=25) Depth of excision Superficial 47 5 6.5 (2.2 to 19.2) Down to fascia 29 20 Complications Yes (one or more) 36 19 3.5(1.3to9.8) None 40 6 Breakdown of complications Hematoma 34 11 Infection 1 4 Prolonged pain 1 4

Table 21 summarizes the correlation between the patient appraisal and the overall cosmetic outcome. When the results were re-classified into two categories, the patient's appraisal of cosmesis was satisfactory in 81.2% (82 of 101) and parallel to the overall outcome in 74.3%. Ten of the 76 (13.2%) patients in the satisfactory group scored "unsatisfactory" and 16 of 25 (64.0%) patients in the unsatisfactory group scored "satisfactory" (p<0.001). 45

Table 21. Relations between patient appraisal and overall cosmetic outcome in 101 wire- guided biopsies of benign breast lesions.

Cosmetic Outcome Patient appraisal Excellent Good Fair Poor Excellent 19 3 1 0 Good 29 15 10 5 Fair 1 8 2 7 Poor 0 1 0 0 6. Discussion

6.1. Subjects and methods

No patients were lost from the series covered in Papers I, II and IV. There was thus no dropout rate to cause any bias in the results. The loss in Paper III was 4% because of missing data in four cases. In Paper V, 18 % of the eligible patients were lost, since four women refused to participate in the study, nine women never attended the cosmetic follow-up visit even after contact by mail or telephone, two patients were unable to attend the cosmetic evaluation because of other diseases, and three women were excluded from analysis because of missing data. Some of the reports concerning FNAB in nonpalpable breast lesions (115,116,118) have been criticized (114,135,137) for excluding the insufficient specimens from their calculations or for treating them as negative results (giving too high sensitivity and specificity values). According to Fornage (137), "the claimed values for specificity of 91%, 95%, 94%, 97%, and 94% drop after correction to 80%, 74%, 78%, 76% and 75%, respectively." The reported insufficient specimen rates of FNAB vary from 0 to 54% (106,114,120,121,258). These rates are very low with US (134,137). There were only three (3%) insufficient specimens in Paper I, which were all included in the calculations. In only part of these previous studies was the correlation of cytology confirmed in all cases by means of histological samples (117), while in the others (105,116,132) the confirmation was mostly based on follow-up times of various length, which is not an equally reliable method as open surgical biopsy (119). In the previous reports of US-FNAB including palpable lesions (29,95), more than half of the final diagnoses have been based on clinical follow-up instead of open surgical biopsies, which is why the accuracy values may be too high. The number of cases in the retrospective series in Paper II was relatively low. Howe- ver, it included all breast operations for persistent unilateral nipple discharge from a single nonlactiferous duct without a palpable mass seen in a large university hospital over a 9-year period. 47

The major problem encountered in the prospective series presented in Paper III origi- nated from the observation that the radiologic orientation of the 17 additional specimens relative to the initial specimen radiography proved to be inaccurate. Therefore, the additional pieces had to be excluded from the analysis of the determinants of positive radiologic margins and from the calculations concerning the correlation between radiologic and histologic margins. The calculations concerning the correlation between radiologic margins and residual disease in re-excisions only included the 54 cases without additional pieces. The obvious benefit of the prospective nature of the study in Paper III was the low rate of missing data. In the retrospective series of Lee et al. (160) concerning the determinants of residual tumor after excisional biopsy of nonpalpable breast cancer, the presence or absence of tumor at the histologic margins of WGB could not be ascertained in 47% of the cases. However, these indeterminate cases were included in their calculations together with those margin-positive for the presence of tumor, which can be criticized (137). The method of assessing lumpectomy margins by touch preparation cytology in Paper IV can be criticized, because an additional slide of the tumor itself as a positive cytologic control was taken from only a small portion of the tumors (12 of 55). This may have complicated the interpretation of cytologic slides, resulting in lower accuracy values. However, the majority (66%) of tumors were nonpalpable and/or small. Transection of the specimen for cytologic slides in such lesions has been discouraged because manipulation may crush the lesion and render it undetectable (3,232). Most reports concerning the evaluation of cosmesis after BCT have been based on the subjective (observer-based) method, which gives an overall score for the degree of breast deformity in three or four discrete cosmetic categories compared with the untreated breast. Subjective evaluation of cosmesis has been criticized in several reports (259-261) for not being always reproducible and not being useful for quantitative measurement of separate factors, since all cosmetic changes are lumped together into one assessment. However, subjective grading of cosmetic results is not without value, and it has been generally used because of a lack of standardized objective (measurement-based) methods. Consensus on cosmesis has exceeded 85% when two-grade instead of four-grade scales have been used (excellent-good vs fair-poor). Experienced observers have attained even more consistent results. (260) Also, the use of a single observer in the postoperative cosmetic evaluation has been claimed to increase validity (255). There are also studies aiming at improved objectivity by quantifying subjective observations. Each type of cosmetic change is evaluated separately by several individually scored specific indices. (257) There have been proposals to use quantitative measurements of nipple displacement and breast contour retraction for objective cosmetic analysis in early breast cancer (259,261,262), but the requirements for special equipment, such as photographs with a bidimensional raster system (262) and a measurement grid (261) with complex calculations makes routine use of such methods difficult (260). Moreover, these methods are not able to identify all the different factors (e.g. disturbing surgical scars) involved in a poor cosmetic outcome (242). In Paper V, cosmesis was evaluated by a subjective method, but instead of placing patients in 3 or 4 discrete cosmetic categories, 6 specific cosmetic indices were individually scored from 0 to 3 in an attempt to quantify the subjective observations. This method was derived from the EORTC breast protocol 10801 rating of cosme- 48 sis (257) with modifications, allowing for the benign nature of the breast lesions. The scale of cosmetic outcome was changed from four grades to two (excellent-good vs fair- poor) for statistical analysis. Almost all evaluations were performed by a single observer.

6.2. US-FNAB in nonpalpable breast lesions (I)

Although the accuracy of cytologic examination in the diagnosis of breast diseases has been widely studied, reports of US-FNAB used solely in nonpalpable breast lesions with pathologic correlation are rare. The accuracy values (sensitivity 84%, specificity 93%) of this study compare well with the corresponding series (sensitivity 88-97 % and specificity 95-99%) (263-264). Previous studies of stereotaxic FNAB in nonpalpable breast lesions have reported similar results (105,116,117,122,132). Half of the malignancies had a maximum US diameter of less than 10 mm, but US-FNAB also proved to be acceptably reliable in such small cancers as reported previously (265). The inherent risk of FNAB is false negativity, which may be attributable to the technical limitations, scant cytologic samples and small size of lesions. All the three false-negative results in this study proved to be sampling errors (two originally interpreted as benign and one of the samples being insufficient) probably due to small size. One of the two specimens originally considered benign was reconsidered to be insufficient at a slide review. All the three patients were referred to WGB because of suggestive findings at mammography. Even though a small risk of false-negative cytologic diagnosis remains, the use of conservative follow-up for cytologically benign lesions instead of WGB is acceptable, if they are mammographically and sonographically benign. Further support is lent to this by the fact that the reported positive predictive value for cancer in periodic mammographic follow-up of cases of probably benign lesions even without fine-needle biopsy has been as low as 0.5% (266). The consequence is a lower cost and less morbidity. One-third of the cytologic results were atypical or indeterminate, and a high proportion (266) can be explained by patient selection (atypical cytologic finding was a definite indi- cation for wire-guided biopsy). The proportion of cancers in the atypical cytologic samples was 24%, which is within the range of previous reports (264,267,268). False-positive results in this study were very rare. The one false-positive result proved to be an interpretative error. The tumor was a fibroadenoma, which is known to cause interpretative difficulties in cytologic samples (269,270). Owing to its very low false- positive rate, a conclusive cytologic diagnosis of malignancy can help in operative planning, thereby reducing the need for two-step operations. Low cost, real-time monitoring of the needle during both insertion and harvesting, reliability and rapidity have made US the preferred guidance method for FNAB of nonpalpable breast lesions that can be visualized sonographically, while the other methods are secondary. To date, however, core- cutting needle biopsy techniques providing sufficient tissue for histologic analysis with a lower rate of insufficient samples have partially replaced FNAB (110,122). 49 6.3. Preoperative methylene blue staining (II)

Malignant neoplasms made up a small minority (7%) and papillary lesions the great majority (intraductal papillomatosis 43%, intraductal papilloma 30%) of the histologic findings in operations performed for nipple discharge in this study, as in many previous reports (29,30,139-142,271). The success rate (73.3%) in preoperative methylene blue dye staining of galactographically suspicious lesions for microdochectomy was only fair. Grillo et al. (271) reported a higher success rate (83.9 %), but both bilateral and palpable lesions were included in their series. Choudhury et al. (149) reported a series of 12 patients with bloody nipple discharge, whose secreting duct was successfully stained peroperatively with methylene blue dye, but none of these patients had had preoperative galactography. They felt that all patients with bloody nipple discharge warrant a biopsy. The mean actual incidence of breast cancer in patients with bloody and serous disharge has been 13% and 7%, respectively (142). As the preoperative methylene blue staining was performed 2-3 weeks after the diagnostic galactography and in many cases by a different operator, it was not always possible to identify the same duct that had been originally visualized. This led to a failure of the staining procedure and to more extensive excision of breast tissue (up to quadrantectomy) in 26.7% (8 of 30 cases), because the peroperative localization had to be based on the clinical and previous galactographic findings. The success rate could easily be raised close to 100%, if all patients with nipple discharge could be scheduled for potential microdochectomy on the day when the primary galactography (and subsequent methylene blue staining) is performed. However, this is not always possible, because most of the diagnostic galactographic examinations are performed outside the surgical unit and only part of them reveal suspicious findings. Thus, referring all women with nipple discharge to a surgical unit is unlikely to be cost-effective.

6.4. WGB for nonpalpable breast cancer (III)

To date, the primary role of WGB has been diagnostic. Its use as a definitive therapeutic procedure is debatable (162). However, no re-excision was needed in a quarter of the present cases, as expected (156,157,161). The results of preoperative needle biopsy were not a topic of this study, but it is most probable that a definitive preoperative diagnosis of malignancy would have increased the proportion of one-step procedures. According to specimen radiographic control, only large mammographic lesions were at a significant risk for radiologically incomplete removal. During WGB, it is more difficult to maintain a 0.5-1 cm margin through a wider resection surface with a mammographically large than small lesion. However, even two-view specimen radiography has limitations, because radiologically apparently clear margins may be found involved if an appropriate tangential view is obtained (159). This problem could only be avoided with a three-dimensional technique, not available with the current imaging equipment. The magnification used in this series most probably improved the accuracy of specimen imaging. There are no previous series addressing the determinants of positive radiologic margins. 50

The mean histologic margin was 5 mm smaller than the radiologic one, as expected (84,156,159). Accurate assessment of histologic margins is difficult (217), and the clinical significance of microscopically positive margins after lumpectomy is debatable (151,157,255), but they usually mandate re-excision of the biopsy site in order to limit local recurrences (272). Histopathological analysis of the present series revealed a 41% incidence of positive microscopic margins. The wide range of this incidence (26-84%) in the literature (151,159,160,233) is probably associated with the great variation in the relative extent of the biopsy procedures in different series (207). In the present study, superficial excision, large pathologic size and multifocality of the lesion proved to be significantly associated with positive histologic margins in multivariate analysis. Extension of the excision lines down to the fascia ensures a minimum risk of involved histologic margins under the index tumor. Larger tumors are, understandably, more difficult to be three- dimensionally completely excised than smaller ones. Tumor foci in the vicinity of the primary tumor may be mammographically undetectable or underestimated (84,100) and therefore prone to be transected by a surgical biopsy incision. Moreover, large (> 2 cm) tumors have been related to a higher rate of surrounding multifocality (241). Multifocal tumors naturally carried a significant risk for residual cancer after WGB in this series. Comparison of the present results with other studies is limited by the differences in the research protocol. The differences in the correlations between specimen radiographic and histologic margins and residual disease in re-excisions compared to the previous retrospective series (159,160) can be addressed to the larger relative extent of the biopsies made in the present prospective series (Table 22). The subsequent lower incidence of involved radiologic and histologic margins reduced the true-positive rate, giving lower sensitivity and positive predictive values. The corresponding increase of true-negative cases resulted in higher specificity and negative predictive values. However, the specificity in the series of Gra- ham et al. (159) was higher, because the mammographic underestimation of the histologic lesion (84,156) combined with their minimal volume biopsy technique resulted in a very low rate of false-positive cases (1 of 119). In the small series of Aitken et al. (273), including palpable cancers, the differences were parallel because of the larger relative extent of biopsy (Table 22). Thus, the accuracy of specimen radiography in defining the adequacy of surgical excision in WGB is clearly dependent on the relative extent of biopsy. The problem is further complicated by the known radiologic underestimation of the histologic extent of the tumor (84,156). In the recent retrospective seriesI of Kollias et al. (212), including mainly palpable cancers, specimen radiography had a specificity of 87% and a positive predictive value of 79% with positive histologic margins (< 3 mm). However, their accuracy values are difficult to compare with the present results, because their positive radiologic margin was defined clearly differently (< 1cm). 51

Taulukko 22. Impact of relative extent of excision on accuracy values between positive specimen radiographic and histologic margins and residual disease in re-excisions.

Radiologic margins Histologic margins Radiologic vs histologic margins vs residual disease vs residual disease (%) (%) (%) Graham et Present Aitken et al Lee et al Present Present Lee et al al (8) study (21)* (9) study study Excision margin (cm) 0 0.5-1 1 0 0.5-1 0 0.5-1 Sensitivity 62 38 11 49 27 93 85 Specificity 95 81 75 77 79 18 59 Positive predictive value 98 58 11 79 33 55 39 Negative predictive value 32 65 75 56 73 67 92 Overall diagnostic accuracy 67 63 61 62 64 58 65

*includes palpable cancers

The histologic margin status can be misleading in predicting residual disease in re- excisions, as reported previously in retrospective series (44,193,211) including palpable invasive breast cancers. This is probably one of the reasons for the contradictory reports concerning the prognostic importance of surgical margins for local control after BSC (195-198,200-205,208,209). The amount of residual cancer can be too small to be detected by standard histologic examination, resulting in negative residual cancer after positive biopsy margins. On the other hand, residual cancer found after negative biopsy margins can be explained by inadequate microscopic examination of margins or multifocality (217). Previous studies of multifocality, including also palpable breast cancers, have reported a 4-20% incidence of residual disease after tumor resection with 4-2 cm margins (240,241).

6.5. Margins and residual tumor after lumpectomy (IV)

Compared to permanent histologic sections, the sensitivity, specificity and overall diagnostic accuracy of touch preparation cytology reported by Cox et al. (229) were 100 %, 96.6 % and 97.3 %, respectively. However, the corresponding figures in the present study were clearly lower. The reasons are not obvious, but an additional cytologic slide of the tumor itself as a positive control was not systematically used and the experience of the cytopathologist was more limited in interpreting touch preparation cytologic slides. In this study, the only significant factors predicting positive histologic margins were large pathologic tumor size and the presence of ductal carcinoma in situ. One previous study noted a correlation between tumor size and margin involvement (206), but no such correlation was seen in another study (207). Since the tumors with positive margins were larger than the ones with negative margins, the sizes of the former must have been underestimated either clinically or mammographically. All but one of the margin-positive can- 52 cers were at least partially noninvasive. As intraductal cancer usually does not induce the kind of desmoplastic reaction typical of invasive cancers, it cannot be palpated as different from the normal surrounding breast tissue during surgery. Often, the only manifesta- tion of DCIS is a cluster of microcalcifications at mammography (85). It is essentially important to carefully assess the size and distribution of the calcifications on preoperative mammograms and specimen radiographs in order to avoid incomplete removal of the tumor. However, not all intraductal tumors are seen at mammography, and a tumor-free margin at specimen radiography is not a good predictor of histologically uninvolved margins (159). Likewise, in situ cancer may be impossible to detect at specimen radiography, if there are no associated microcalcifications. Only 20.6 % (7 of 34) of the re-excision specimens in this study contained residual carcinoma. Notably, the histologic or cytologic margin status was not a reliable predictor of residual cancer in re-excision specimens, as also reported previously (87,193,211,213). One reason for negative residual cancer after positive lumpectomy margins may be that the amount of residual cancer is too small to be detected by a standard histologic examination. Multifocality, on the other hand, explains the residual cancers found after negative lumpectomy margins (240,241). According to Morimoto et al. (241), lumpectomies to remove T1 tumors with margins of 2 cm would leave residual cancer foci in the breast in about 20% of cases. In the present study, multifocality was also the most important predicting factor. Another predictor of residual disease was nonpalpability of the index tumor. The most probable reason for this is that nonpalpable cancers were more often multifocal and noninvasive than palpable ones. Nonpalpable breast tumors have also earlier been reported to involve a higher rate of postbiopsy residual disease (274). Moreover, microcalcifications in preoperative mammograms predicted residual disease. This was not unexpected, as microcalcifications are an important sign of DCIS, the size of which is often underestimated at mammography (84). All residual tumors contained noninvasive cancer.

6.6. Cosmesis (V)

The cosmetic outcome of WGB for nonpalpable breast lesion in this study was excellent or good in 75.5% of the cases. Corresponding figures of 49% to 89% have been reported after BCT, but any comparison with the present results is difficult. The adverse impact of irradiation on the overall cosmetic outcome is almost impossible to evaluate. It is notable (224,242-246,250,275,276), however, and such changes were absent in our series. Cosmesis was evaluated 6 months after breast biopsy, because the surgical changes alone in the breast tend to be evident within months after lumpectomy (185), and human scar tissue maintains a relatively high level of collagen synthesis for only up to 6 months (251), though remodelling of the wound continues for 12 months or more (277). Notably, cosmetic scores after BCT deteriorate for the first 30-60 months because of postradiation changes (243,244,252). More superficial excisions resulted in better cosmesis compared to biopsies extending down to the fascia. Thus, a rim of normal breast tissue should preferably be saved on the fascia, unless complete excision cannot be ensured otherwise. In BCT, deep excisions 53 often result in retraction due to scar tissue, which is further aggravated by radiation-induced fibrosis (253). The adverse impact of postoperative complications is logical and expected, especially with regard to infections. Notably, large specimen volume had no statistically significant adverse impact on cosmetic outcome, contrary to BCT (224,242- 249,252,254). Large specimen volumes in BCT are probably related to a large tumor size and the doses of irradiation administered. However, resection volumes in BCT are significantly larger than the mean volume of 31 cm3 in the present study. Other factors, such as age, breast size, body mass index, and location of the lesion, had no impact on the esthetic outcome. In BCT, some of these factors have been related to a lower cosmetic grading (242,243,245,246,250,278), but the results have been controversial. There were too few cases of radial skin incision, skin excised with the specimen, and intraoperative blood loss of more than 50 mL to allow any definitive conclusions. The patient's appraisal of the cosmetic outcome agreed with the overall cosmetic outcome in 74% of the cases in the satisfactory vs. unsatisfactory rating categories. The woman's own evaluation was generally more positive than the overall cosmetic outcome. This has also been the case in studies concerning cosmesis after BCT (221,243,246,250). However, there were ten patients (13%) who considered their cosmetic outcome unsatisfactory rather than satisfactory, contrary to the overall outcome, a feature which is uncommon after BCT (250). Nevertheless, this was expected, as women with a benign diagnosis have been reported to be twice as likely to be discontent with the new appearance of the breast than those with breast cancer (279).

6.7. General discussion

The results of this study emphasize the paramount importance of a correct preoperative diagnosis and close co-operation between the radiologist, surgeon and breast pathologist in the surgical treatment of suspicious nonpalpable breast lesions and early breast cancer. Owing to the high accuracy of US-FNAB, a conclusive cytologic diagnosis of malignancy crucially facilitates thorough pretreatment planning and discussion with the patient. With an established diagnosis of nonpalpable breast cancer, WGB may be increasingly employed as a definitive therapeutic procedure. The extent of excision in nonpalpable and early breast cancer can be more easily individualized between adequate local control and cosmesis. According to the European Society of Surgical Oncology and the European Breast cancer Working Group, 90% of palpable and the majority of nonpalpable breast cancers should be diagnosed preoperatively by fine-needle cytology or core needle histology (111,223). The results also underscore the importance of the knowledge of the relative extent of surgical procedures in the evaluation of the reported treatment results in BCT. To date, however, the major unresolved issues regarding local treatment techniques in BCT include the optimal method of assessing lumpectomy margins and the optimal extent of excision. The results of this study suggest that the relative extent of excision should be adjusted to the size of the primary lesion. To achieve grossly and microscopically uninvolved margins, small (< 1.5 cm) primary tumors should be excised with a normal tissue margin of approximately 1 cm. In large (>1.5cm) lesions, excision extending down to the 54 fascia should include a normal tissue margin of approximately 2 cm. If positive histologic margins or multifocality are encountered in the histologic examination of the specimen, re-excision of the index tumor site should be considered due to the increased risk of residual cancer. In the near future, the combined use of preoperative imaging modalities, especially image-guided needle biopsy, may replace the vast majority of diagnostic WGBs in evaluating patients with nonpalpable breast lesions. The result is lower cost, less morbidity and better cosmesis, even though the cosmetic results after WGB of benign breast lesions can be considered satisfactory. It is, however, not possible to replace all surgical excisions of benign breast lesions with needle biopsy or follow-up. A woman with a breast lesion may prefer operation instead of follow-up. Most nonpalpable intraductal lesions with nipple discharge remain nonspecific at mammography and galactography. They are difficult or impossible to exa- mine with needle biopsy. Even in these cases, accurate preoparative diagnosis and localization of the lesion are crucial for the surgeon to enable accurate excision of optimal extent with a good cosmetic outcome. 7. Conclusions

1. US-FNAB is a useful tool in evaluating nonpalpable suspicious breast lesions in normal clinical practice. 2. Preoperative methylene blue staining crucially facilitates selective minimal-volume microdochectomy in three quarters of nonpalpable galactographically suspicious breast lesions. An interval between primary galactography and later methylene blue staining may be associated with a failure to identify the originally visualized duct in approximately one quarter of the cases. 3. To obtain free margins in WGB for nonpalpable breast cancer and in lumpectomy for early breast cancer, mammographically and pathologically large lesions should be removed with wider excisions, preferably extending down to the pectoralis fascia. However, radiologic and histologic margins of the specimen may be misleading. The biopsy site of multifocal tumors should be re-excised because of the considerable risk of residual disease. 4. Touch preparation cytology cannot be recommended as a method of assessing lumpectomy margins for early breast cancer. 5. The cosmetic outcome after WGB of benign breast lesions is excellent or good in 75 % of cases. Excisions extending down to the pectoralis fascia and complications endanger the esthetic outcome. 8. References

1. Anttila A, Auvinen A, Hakama M, Hakulinen T, Koivusalo M, Lehtonen M, Luosta- rinen T, Määttänen L, Pukkala E, Sankila E, Söderman B, Teppo L & Voutilainen E (1997) Cancer incidence in Finland 1995. Finnish Cancer Registry, Helsinki, Fin- land. 2. Broeders MJ & Verbeek AL (1997) Breast cancer epidemiology and risk factors. Q J Nucl Med 41: 179-188. 3. Suomen Rintasyöpäryhmä ry (1999) Rintasyövän valtakunnallinen diagnostiikka ja hoitosuositus. Kuopio: Finnish Breast Cancer Group. 4. Tabar L, Duffy SW, Vitak B, Chen HH & Prevost TC (1999) The natural history of breast carcinoma: what have we learned from screening? Cancer 86: 449-462. 5. Tabar L, Fagerberg G, Duffy SW & Day NE ( 1989) The Swedish two county trial of mammographic screening for breast cancer: recent results and calculation of benefit. J Epidemiol Community Health 43: 107-114. 6. Hendrick RE, Smith RA, Rutledge JH 3rd & Smart CR (1997) Benefit of screening mammography in women aged 40-49: a new meta-analysis of randomized controlled trials. J Natl Cancer Inst Monogr 22: 87-92. 7. Larsson LG, Andersson I, Bjurstam N, Fagerberg G, Frisell J, Tabar L & Nystrom L (1997) Updated overview of the Swedish Randomized Trials on Breast Cancer Screening with Mammography: age group 40-49 at randomization. J Natl Cancer Inst Monogr 22: 57-61. 8. Tabar L, Duffy SW & Krusemo UB (1987) Detection method, tumor size and node metastases in breast cancer diagnosed during a trial of breast cancer screening. Eur J Cancer Clin Oncol 25: 959-962. 9. Letton HA, Mason EM & Ramshaw BJ (1996) Twenty-year review of a breast cancer screening project. Cancer 77: 104-106. 10. Bauer TL, Pandelidis SM & Rhoads JE Jr (1994) Five-year survival of 100 women with carcinoma of the breast diagnosed by screening mammography and needle- localization biopsy. J Am Coll Surg 178: 427-430. 57

11. Tabar L, Fagerberg G, Gad A, Baldetorp L, Holmberg LH, Gröntoft O, Ljungquist U, Lundström B, Månson JC, Eklund G, Day NE & Pettersson F (1985) Reduction in mortality from breast cancer after mass screening with mammography. Randomised trial from the Breast Cancer Screening Working Group of the Swedish National Board of Health and Welfare. Lancet 1: 829-832. 12. Carter CL, Allen C & Henson DE (1989) Relation of tumor size, lymph node status, and survival in 24.740 breast cancer cases. Cancer 63: 181-187. 13. Clay MG, Hislop G, Kan L, Olivotto IA & Burhenne LJ (1994) Screening mammography in the British Columbia 1988-1993. Am J Surg 167: 490-492. 14. Cady BI (1997) Traditional and future management of nonpalpable breast cancer. Am Surg 63: 55-58. 15. Perdue P, Page D, Nellestein M, Salem C, Galbo C & Ghosh B (1992) Early detection of breast carcinoma: a comparison of nonpalpable and palpable lesions. Surgery 111: 656-659. 16. Solin LJ, Fowble BL, Schultz DJ & Goodman RL (1990) The impact of mammography on the patterns of patients referred for definitive breast irradiation. Cancer 65: 1085-1089. 17. Fagerberg G, Baldetorp L, Gröntoft O, Lundström B, Månson JC & Nordenskjöld B (1985) Effects of repeated mammographic screening on breast cancer stage distribution: results from a randomized study of 92, 934 women in a Swedish county. Acta Radiol Oncol 24: 465-473. 18. Tabar L, Fagerberg G, Duffy SW, Day NE, Gad A & Gröntoft O (1992) Update of the Swedish two-county program of mammographic screening for breast cancer. Radiol Clin North Am 30: 187-210. 19. Wong WW, Vijayakumar S & Weichselbaum RR (1992) Prognostic indicators in node-negative early breast cancer. Am J Med 92: 539-548. 20. Winchester DP (1991) Adjuvant therapy for node-negative breast cancer. The use of prognostic factors in selecting patients. Cancer 67: 1741-1743. 21. National Institute of Health Consensus Conference (1991) Treatment of early-stage of breast cancer. JAMA 265: 391-395. 22. Boshaw DL (1997) New instrumentation allows removal of nonpalpable breast lesions. AORN J 66: 296-302. 23. Drukker BH (1997) Breast disease: A primer on diagnosis and management. Int J Fertil 42: 278-287. 24. Haffty BG, Lee C, Philpotts L, Horvath L, Ward B, McKhann & Tocino I (1998) Prognostic significance of mammographic detection in a cohort of conservatively treated breast cancer patients. Cancer J Sci Am 4: 35-40. 25. Silverstein MJ, Gamagami P, Masetti R, Legmann MD, Craig PH & Gierson ED (1997) Results from a multidisciplinary breast center. Surg Oncol Clin North Am 6: 301-314. 26. Devia A, Murray KA & Nelson EW (1997) Stereotactic core needle biopsy and the workup of mammographic breast lesions. Arch Surg 132: 512-515. 27. Yim JH, Barton P, Weber B, Radford D, Levy J, Monsees B, Flanagan F, Norton JA & Doherty GM (1996) Mammographically detected breast cancer. Benefits of stereotactic core versus wire localization biopsy. Ann Surg 223: 688-700. 58

28. Chaudary MA, Millis RR, Path GCD & Hayward JL (1982) Nipple discharge: the diagostic value of testing occult blood. Ann Surg 196: 651-655. 29. Paterok EM, Rosenthal H & Säbel M (1993) Nipple discharge and abnormal galac- togram. Eur J Obstet Gynecol Reprod Biol 50: 227-234. 30. Gulay H, Bora S, Kilicturgay S, Hamaloglu E & Goksel HA (1994) Management of nipple discharge. J Am Coll Surg 178: 471-474. 31. Urban JA & Egeli RA (1978) Non-lactational nipple discharge. Ca-A Cancer J Clin 28: 130-140. 32. Welch M, Durrans D, Gonzalez J, Daya H & Owen AM (1990) Microdochectomy for discharge from a single lactiferous duct. Br J Surg 77: 1213-1214. 33. Fung A, Rayter Z, Fischer C, King DM & Trott P (1990) Preoperative cytology and mammography in patients with single-duct nipple discharge treated by surgery. Br J Surg 77: 1211-1212. 34. Baker KS, Davey DD & Stelling CB (1994) Ductal abnormalities detected with galactography: frequency of adequate excisional biopsy. AJR 162: 821-824. 35. Vizcaino I, Salas D, Vilar JS, Ruiz-Perales F, Herranz C & Ibanez J (1998) Breast cancer screening: first round in the population-based program in Valencia, Spain. Collaborative Group of Readers of the Breast Cancer Screening Program of the Valencia Community. Radiology 206: 253-260. 36. Dean PB & Pamilo M (1999) Screening mammography in Finland: 1.5 million examinations with 97% specificity. Acta Oncol, in press. 37. Alexander HR, Candela FC, Dershaw DD & Kinne DW (1990) Needle-localized mammographic lesions. Results and evolving treatment strategy. Arch Surg 125: 1441-1444. 38. McCreery BR, Frankl G & Frost DB (1991) An analysis of the results of mammographically guided biopsies of the breast. Surg Gynecol Obst 172: 223-226. 39. Thompson WR, Bowen JR & Dorman BA (1991) Mammographic localization and biopsy of nonpalpable breast lesions. Arch Surg 26: 730-734. 40. Jeter DD, Vest GR & Buday SJ (1991) Mammographic guide-wire localization of nonpalpable breast lesions. Am Surg 57: 431-433. 41. Radford DM, Cromack DT, Troop BR, Keller SM & Lopez MJ (1992) Pathology and treatment of impalpable breast lesions. Am J Surg 164: 427-432. 42. Symonds DA, Copeland BE, Drane A, Kaplan GN & Graham RR (1992) Pathologic correlation in mammographically directed breast biopies. Arch Pathol Lab Med 116: 28-31. 43. Fisher B, Redmond C, Poisson R, Margolese R, Wolmark N, Wickerham L, Fisher E, Deutsch M, Caplan R & Pilch Y (1989) Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without radiation in the treatment of breast cancer. N Engl J Med 320: 822-828. 44. Fisher B, Anderson S,Redmond CK, Wolmark N, Wickerham DL & Cronin W (1995) Reanalysis and results after 12 years of follow-up in a randomized clinical trial comparing total mastectomy with lumpectomy with or without irradiation in the treatment of breast cancer. N Engl J Med 30: 1456-1461. 45. Fisher B, Redmond C & Others for the National Surgical Adjuvant Breast and Bowel Project (1992) Lumpectomy for breast cancer: an update of the NSABP experience. J Natl Cancer Inst Monogr 11: 7-13. 59

46. Mansfield CM, Komarnicky LT, Schwartz GF, Rosenberg AL, Krishnan L, Jewell WR, Rosato FE, Moses ML, Haghbin M & Taylor J (1995) Ten-year results in 1070 patients with stages I and II breast cancer treated by conservative surgery and radiation therapy. Cancer 75: 2328-2336. 47. Rubin E, Dempsey PJ, Pile NS, Bernreuter WK, Urist MM, Shumate CR & Maddox WA (1995) Needle-localization biopsy of the breast: impact of a selective core needle biopsy program on yield. Radiology 195: 627-631. 48. Ewertz M (1996) Epidemiology of breast cancer: the nordic contribution. Eur J Surg 162: 97-99. 49. Pisani P, Parkin DM & Ferlay J (1993) Estimates of the worlwide mortality from eighteen major cancers in 1985: implications for prevention and projections of future burden. Int J Cancer 55: 891-903. 50. Eskelinen M (1992) Rintasyövän kliininen diagnostiikka. Duodecim 108: 1859- 1868. 51. Gilbertsen VA (1974) The earlier detection of breast cancer. Semin Oncol 1: 87-89. 52. Hicks MJ, Davis JR & Layton JM (1979) Sensitivity of mammography and physical examination of the breast for detecting breast cancer. JAMA 242: 2080-2083. 53. Bender HG, Schnurch HG & Beck L (1988) Breast cancer detection: age-related significance of findings on physical exam and mammography. Gyn Oncol 31: 166-172. 54. Feig SA, Shaber GS, Schwartz GF, Patchefsky A, Libshitz HI, Edeiken J, Nerlinger R, Curley RF & Wallace JD (1977) Thermography, mammography and clinical examination in breast cancer screening: review of 16,000 studies. Radiology 122: 123- 127. 55. McKenna RJ Sr, Greene P, Winchester DP, Baines CJ, Foster RS, Champion V & O'Malley MS (1992) Breast self-examination and breast physical examination. Can- cer 69: 2003-2004. 56. Koibuchi Y, Ino Y, Takei H, Maemura M, Horiguchi J, Yokoe T & Morishita Y (1998) The effect of mass screening by physical examination combined with regular breast self-examination on clinical stage and course of Japanese women with breast cancer. Oncology Reports 5: 151-155. 57. Semiglazov VF, Moiseenko VM, Protsenko SA, Bavli IL, Orlov AA, Ivanova OA, Barash NI, Chagunava OL, Golubeva OM, Migmanova NS, Seleznev IK, Popova RT, Diatchenko OT, Kozhevnikov SI, Aleksandrova GI, Sanchakova AV, Kharikova RS, Liubomirova NK, Ivanova GV, Azeev VF & Chuprakova IS (1996) Preliminary results of the Russia (St. Petersburg)/WHO program for the evaluation of the effectiveness of breast self-examination. Vopr Onkol 42: 49-55. 58. Holmberg L, Ekbom A, Calle E, Mokdad A & Byers T (1997) Breast cancer mortality in relation to self-reported use of breast self-examination. A cohort study of 450,000 women. Breast Cancer Res Treat 43: 137-140. 59. Winchester DP (1992) Physical examination of the breast. Cancer 69: 1947-1949. 60. Mendelson EB (1992) Evaluation of the postoperative breast. Radiol Clin North Am 30: 107-138. 61. Egan RL (1960) Experience with mammography in a tumor institute: evaluation of 1,000 studies. Radiology 75: 894-900. 62. Tabar L & Dean PB (1983) Teaching Atlas of Mammography. Georg Thieme Verlag, Stuttgart, New York. 60

63. Evans WP (1995) Breast masses: appropriate evaluation. Radiol Clin North Am 33: 1085-1108. 64. Sickles EA (1991) Periodic mammographic follow-up of probably benign lesions: results in 3184 consecutive cases. Radiology 179: 463-468. 65. McGreevy JM & Loftus TJ (1998) Outcomes evaluation for operative and nonopera- tive management of the abnormal mammogram. Am J Surg 175: 69-72. 66. Morrow M, Schmidt R, Cregger B, Hassett C & Cox S (1994) Preoperative evaluation of abnormal mammographic findings to avoid unnecessary breast biopsies. Arch Surg 129: 1091-1096. 67. Kalisher L (1979) Factors influencing false negative rates in xeromammography. Radiology 133: 297-301. 68. Wolfe JN (1964) Mammography: report on its use in women with breast abnormal and normal on physical examination. Radiology 83: 244-254. 69. Wallis MG, Walsh MT & Lee JR (1991) A review of false negative mammography in a symptomatic population. Clin Radiol 44: 13-15. 70. Martin JE, Moskowitz M & Milbrath JR (1979) Breast cancers missed by mammography. AJR 132: 737-739. 71. Cahill CJ, Boulter, PS, Gibbs NM & Price JL (1981) Features of mammographically negative breast tumours. Br J Surg 68: 882-884. 72. Young JO, Sadowsky NL, Young JW & Herman L (1986) Mammography of women with suspicious breast lumps. Arch Surg 121: 807-809. 73. Locker AP, Stickland V, Manhire AR, Caseldine J & Blamey RW (1989) Mammo- graphy in symptomatic breast disease. Lancet 1: 887-889. 74. Coveney EC, Geraghty JG, O'Laoide R, Hourihane JB & O'Higgins NJ (1994) Reasons underlying negative mammography in patients with palpable breast cancer. Clin Radiol 49: 123-125. 75. Edeiken S (1988) Mammography and palpable cancer of the breast. Cancer 61: 263- 265. 76. Nilhoff PH & Sheiner NM (1981) False negative mammograms in patients with breast cancer. Can J Surg 24: 50-52. 77. Egeli RA & Urban JA (1979) Mammography in symptomatic women 50 years of age and under and those over 50. Cancer 43: 878-882. 78. Mann BD, Giuliano AE, Bassett LW, Barber MS, Hallauer W & Morton DL (1983) Delayed diagnosis of breast cancer as a result of normal mammograms. Arch Surg 118: 23-24. 79. Kaufman Z, Garstin WI, Hayes R, Michell MJP, Baum M (1991) The mammographic parenchymal patterns of women on hormonal replacement therapy. Clin Radiol 43: 389-392. 80. Rosenberg RD, Hunt WC, Williamson MR, Gilliland FD, Wiest PW, Kelsey CA, Key CR & Linver MN (1998) Effects of age, breast density, ethnicity, and estrogen replacement therapy on screening mammographic sensitivity and cancer stage at diagnosis: review of 183,134 screening mammograms in Albuquerque, New Mexico. Radiology 209: 511-518. 81. Devitt JE (1985) Management of nipple discharge by clinical findings. Am J Surg 149: 789-792. 61

82. Simonetti G, Cossu E, Montanaro M, Caschili C & Giuliani V (1998) What's new in mammography. Eur J Radiol 27: 234-241. 83. Morrow M, Schmidt RA & Bucci C (1998) Breast conservation for mammographically occult carcinoma. Ann Surg 227: 502-506. 84. Holland R, Hendricks JHCL, Verbeek ALM, Mravunic M & Schuurmans-Stekhoven JH (1990) Extent, distribution, and mammographic/histological correlations of breast ductal carcinoma in situ. Lancet 335: 519-522. 85. Sickles E (1986) Breast calcifications: mammographic evaluation. Radiology 160:289-293. 86. Monsees BS (1995) Evaluation of breast microcalcifications. Radiol Clin North Am 33: 1109-1119. 87. Noguchi M, Katev N & Miyazaki I (1996) Clinical and histologic determinations of adequate breast resection in breast-conserving surgery. Int Surg 81: 309-315. 88. Jackson VP (1995) The current role of ultrasonography in breast imaging. Radiol Clin North Am 33: 1161-1170. 89. Mendelson EB & Tobin CE (1995) Critical pathways in using breast US. Radio- Graphics 15: 935-945. 90. Sohn C, Blohmer J-U & Hamper UM (1999) Breast Ultrasound - A Systematic Approach to Technique and Image Interpretation. Thieme, Stuttgart, New York. 91. Stavros AT, Thickman D, Rapp CL, Dennis MA, Parker SH & Sisney GA (1995) Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. Radiology 196: 123-134. 92. Reinikainen HT, Rissanen TJ, Piippo UK & Päivänsalo MJ (1999) Contribution of ultrasonography and fine needle aspiration cytology to the differential diagnosis of nonpalpable solid breast lesions. Acta Radiol (in press). 93. DiFronzo LA & O'Connell TX (1996) Breast cancer in pregnancy and lactation. Surg Clin North Am 76: 267-278. 94. Gordon PB, Goldenberg SL & Chan NHL (1993) Solid breast lesions: diagnosis with US-guided fine-needle aspiration biopsy. Radiology 189: 573-580. 95. Fornage BD, Faroux MJ & Simatos A (1987) Breast masses: US-guided fine-needle aspiration biopsy. Radiology 162: 409-414. 96. Hieken TJ & Velasco JM (1998) A prospective analysis of office-based breast ultrasound. Arch Surg 133: 504-508. 97. Rissanen T, Pamilo M & Suramo I (1998) Ultrasonography as a guidance method in the evaluation of mammographically detected nonpalpable breast lesions of sus- pected malignancy. Acta Radiol 39: 292-297. 98. Fischer U, Kopka L & Grabbe E (1998) Magnetic resonance guided localization and biopsy of suspicious breast lesions. Top Magn Reson Imaging 9: 44-59. 99. Muller-Schimpfle M, Stoll P, Stern W, Kurz S, Dammann F & Claussen CD (1997) Do mammography, sonography, and MR mammography have a diagnostic benefit compared with mammography and sonography? AJR 168: 323-329. 100. Boetes C, Mus RDM, Holland R, Barentsz JO, Strijk SP, Wobbes T, Hendriks JHCL & Ruys SHJ (1995) Breast tumors: comparative accuracy of MR imaging relative to mammography and US for demonstrating extent. Radiology 197: 743-747. 62

101. Muller R-D, Barkhausen J, Sauerwein W & Langer R (1998) Assessment of local recurrence after breast-conserving therapy with MRI. J Comput Assist Tomogr 22: 408-412. 102. Morris EA, Schwartz LH, Dershaw DD, Van Zee KJ, Abramson AF & Liberman L (1997) MR imaging of the breast in patients with occult primary breast carcinoma. Radiology 205: 437-440. 103. Palmedo H, Grunwald F, Bender H, Schomburg A, Mallman P, Krebs D & Biersack HJ (1996) Scintimammography with technetium-99m methoxyisobutylisonitrile: comparison with mammography and magnetic resonance imaging. Eur J Nucl Med 23: 940-946. 104. Fuhrman GM, Cederbom GJ, Bolton JS, King A, Duncan JL, Champaign JL, Smet- herman DH, Kuske RR & McKinnon WM (1998) Image-guided core needle breast biopsy is an accurate technique to evaluate patients with nonpalpable imaging abnormalities. Ann Surg 227: 932-939. 105. Azavedo E, Auer G & Svane G (1989) Stereotactic fine-needle biopsy in 2594 mammographically detected nonpalpable lesions. Lancet 1: 1033-1036. 106. Dowlatshahi K,Yaremko ML, Kluskens LF & Jokich PM (1991) Nonpalpable lesions: findings of stereotaxic needle-core biopsy and fine-needle aspiration cytology. Radiology 181: 745-750. 107. Smith DN, Christian R & Meyer (1997) Large-core needle biopsy nonpalpable breast cancers. The impact on subsequent surgical excisions. Arch Surg 132: 256- 260. 108. Seoudi H, Mortier J, Basile R & Curletti E (1998) Stereotactic core needle biopsy of nonpalpable breast lesions: initial experience with a promising technique. Arch Surg 133: 366-372. 109. Sadowsky NL, Semine AA, Levine E & Kelly E (1997) Ultrasonographic guidance for needle biopsy of breast lesions. Surg Oncol Clin N Am 6: 265-284. 110. Head JF, Haynes AE, Elliott MC & Elliot RL (1996) Stereotaxic localization and core needle biopsy of nonpalpable breast lesions: two-year follow-up of a prospective study. Am Surg 62: 1018-1023. 111. O'Higgins N, Linos DA, Blichert-Toft M, Cataliotti L, de Wolf C, Rochard F, Rut- gers EJ, Roberts PJ, Mattheiem W, da Silva MA, Holmberg L, Schulz KD, Smola MG & Mansel RE (1998) European guidelines for quality assurance in the surgical management of mammographically detected lesions. European Breast Cancer Wor- king Group. Ann Chir Gynaecol 87: 110-112. 112. Kaufman CS, Delbecq R & Jacobson L (1998) Excising the reexcision: stereotactic core-needle biopsy decreases need for reexcision of breast cancer. World J Surg 22: 1023-1028. 113. Whitten TM, Wallace TW, Bird RE & Turk PS (1997) Image-guided core biopsy has advantages over needle localization biopsy for the diagnosis of nonpalpable breast cancer. Am Surg 63 1072- 1077. 114. Jackson VP (1992) The status of mammographically guided fine-needle aspirations biopsy of nonpalpable breast lesions. Radiol Clin North Am 30: 155- 166. 115. Helvie MA, Baker DE, Adler DD, Andersson I, Naylor B & Buckwalter KA (1990) Radiographically guided fine-needle aspiration of nonpalpable breast lesions. Radio- logy 174: 657-661. 63

116. Ciatto S, Rosselli Del Turco MR & Bravetti P (1989) Nonpalpable breast lesions: stereotactic fine-needle aspirations cytology. 173: 57-59. 117. Dowlatshahi K, Gent HJ, Schmidt R, Jokich PM, Bibbo M & Sprenger E (1989) Nonpalpable breast tumors: diagnosis with stereotaxic localization and fine-needle aspiration. Radiology 170: 427-433. 118. Bibbo M, Schreiber M, Cajulis R, Keebler CM, Wied GL & Dowlatshahi K (1988) Stereotaxic fine-needle aspiration cytology of clinically occult malignant and prema- lignant breast lesions. Acta Cytol 32: 192-201. 119. Lifrange E, Kridelka F & Colin C (1997) Stereotaxic needle core-biopsy and fine- needle aspiration biopsy in the diagnosis of nonpalpable breast lesions: controversies and future prospects. Eur J Radiol 24: 39-47. 120. Dowlatshahi K, Jokich PM, Schmidt R, Bibbo M & Dawson PJ (1987) Cytologic diagnosis of occult breast lesions using stereotaxic needle aspiration. Arch Surg 122: 1343-1346. 121. Pisano ED, Fajardo LL, Tsimikas J, Sneige N, Frable WJ, Gatsonis CA, Evans WP, Tocino I & McNeil BJ (1998) Rate of insufficient samples for fine-needle aspiration for nonpalpable breast lesions in a multicenter clinical trial: The Radiologic Diag- nostic Oncology Group 5 Study. The RDOG5 investigators. Cancer 82: 679-688. 122. Elvecroq EL, Lechner MC & Nelson MT (1993) Nonpalpable breast lesions: correlation of stereotaxic large-core needle biopsy and surgical biopsy results. Radiology 188: 453-455. 123. Evans WP (1992) Fine-needle aspiration cytology and core biopsy of nonpalpable breast lesions. Curr Opin Radiol 4: 130-138. 124. Pettine S, Place R, Babu S, Wiiliard W, Kim D & Carter P (1996) Stereotactic breast biopsy is accurate, minimally invasive, and cost effective. Am J Surg 171: 474-476. 125. Pitre B, Baron PL, Baron LF, O'Brien PH & Cole DJ (1997) Stereotactic core biopsy of the breast: results of one-year follow-up of 101 patients. Am Surg 63: 1124-1127. 126. Acheson MB, Patton RG, Howisey R, Lane RF & Morgan A (1997) Histologic correlation of image-guided core biopsy of nonpalpable breast lesions. Arch Surg 132: 815-818. 127. Ewans WP (1992) Fine-needle aspiration cytology and core biopsy of nonpalpable breast lesions. Current Opinion in Radiology 4: 130-138. 128. Zannis VJ & Aliano KM (1998) The evolving practice pattern of the breast surgeon with disappearance of open biopsy for nonpalpable lesions. Am J Surg 176: 525- 528. 129. Liberman L, Dershaw DD, Rosen PP, Giess CS, Cohen MA, Abramson AF & Hann LE (1995) Stereotaxic core biopsy of breast carcinoma: accuracy at predicting invasion. Radiology 194: 379-381. 130. Parker SH, Stavros T & Dennis MA (1995) Needle biopsy techniques. Radiol Clin North Am 33: 1171-1186. 131. Burbank F (1997) Stereotactic breast biopsy of atypical ductal hyperplasia and ductal carcinoma in situ lesions: improved accuracy with directional, vacuum-assisted biopsy. Radiology 202: 843-847. 132. Oliver DJ, Frayne JR & Sterret G (1992) Stereotactic fine-needle biopsy of the breast. Aust N Z J Surg 62: 463-467. 64

133. Gordon PB, Goldenberg SL & Chan NHL (1993) Solid breast lesions: diagnosis with US-guided fine-needle aspiration biopsy. Radiology 189:573-580. 134. Okamoto H, Ogawara T, Inoue S, Kobayashi K, Sekikawa T & Matsumoto Y (1998) Clinical management of nonpalpable or small breast masses by fine-needle aspiration biopsy (FNAB) under ultrasound guidance. J Surg Oncol 67: 246-250. 135. Fornage BD (1991) Percutaneus biopsies of the breast: state of the art. Cardiovasc Interv Radiol 14: 29-39. 136. Fornage BD, Coan JD & David CL (1992) Ultrasound-guided needle biopsy of the breast and other interventional procedures. Radiol Clin North Am 30: 167-185. 137. Fornage BD (1990) Guided fine-needle aspiration biopsy of nonpalpable breast lesions: calculation of accuracy values (letter). Radiology 177: 884-885. 138. Liberman L, Feng TL, Dershaw DD, Morris EA & Abramson AF (1998) US-guided core breast biopsy: use and cost-effectiveness. Radiology 208: 717-723. 139. Funderburk WW & Syphax B (1969) Evaluation of nipple discharge in benign and malignant diseases. Cancer 24: 1290-1296. 140. Hou MF, Huang CJ, Huang YS, Huang TJ, Chan HM, Wang JY, Liu GC & Wu DK (1998) Evaluation of galactography for nipple discharge. Clin Imaging 22: 89-94. 141. Rongione AJ, Evans BD, Kling KM & McFadden DW (1996) Ductography is a useful technique in evaluation of abnormal nipple discharge. Am Surg 62: 785-788. 142. Tabar L, Dean PB, Pentek Z (1983) Galactography: the diagnostic procedure of choice in nipple discharge. Radiology 149: 31-38. 143. Ito Y, Tamaki Y, Nakano Y, Kobayashi T, Takeda T, Wakasugi E, Miyashiro I, Kom- oike Y, Miyazaki M, Nakayama T, Kano T & Monden M (1997) Nonpalpable breast cancer with nipple discharge: how should it be treated? Anticancer Res 17: 791-794. 144. Fiorica J (1994) Nipple discharge. Obstet Gynecol Clin North Am 21: 453-460. 145. Kåresen R, Skaane P & Skjorten F (1993) A simple method for excision of intraductal changes in the breast. Tidsskr Nor Laegeforen 113: 1233-1235. 146. Van Zee KJ, Ortega Perez G, Minnard E & Cohen MA (1998) Preoperative galactography increases the diagnostic yield of major duct excision for nipple discharge. Cancer 82: 1874-1880. 147. Locker AP, Galea MH, Ellis IO, Holliday HW, Elston CW & Blamey RW (1988) Microdochectomy for single-duct discharge from the nipple. Br J Surg 75: 700-701. 148. Berna JD, Madrigal M, Guirao J, Arcas I & Gomez S (1990) Microdochectomy: the precise identification of the suspicious duct. Br J Surg 77: 1217-1218. 149. Choudhury A, Wengert PA & Smith JS (1989) A new surgical localization technique for biopsy in patients with nipple discharge. Arch Surg 124: 874-875. 150. Markopoulos C, Kakisis J, Kouskos S, Kontzoglou K, Koufopoulos K & Gogas J (1999) Management of nonpalpable, mammographically detectable breast lesions. World J Surg 23: 434-438. 151. Mokbel K, Ahmed M, Nash A & Sacks N (1995) Reexcision operations in nonpalpable breast cancer. J Surg Oncol 58: 225-228. 152. Ahovuo J, Kinnunen J, von Smitten K & Edgren J (1990) Wire-guided biopsies of nonpalpable breast lesions (1990) Ann Chir Gynaecol 79: 82-84. 153. Aitken RJ, MacDonald HL, Kirkpatrick AE, Anderson TJ, Chetty U & Forrest APM (1990) Outcome of surgery for nonpalpable mammographic abnormalities. Br J Surg 77: 673-676. 65

154. Howe JR, Monsees B, Destouet J, Seib J, Dehner LP & Kraybill WG (1995) Needle localization breast biopsy: a model for multidisciplinary quality assurance. J Surg Oncol 58: 233-239. 155. Hall JA, Murphy DC, Hall BR & Hall KA (1998) Open surgical biopsy for nonpalpable mammographic abnormalities: still an option compared with core needle biopsy. Am J Obstet Gynecol 178: 1245-12450. 156. Hastrich DJ, Dunn JM, Armstrong JS, Davies JD, Davies ZD, Webb AJ & Farndon JR (1992) Diagnostic and therapeutic aspects of fine-wire localization biopsy of impalpable breast cancer. Br J Surg 79: 1038-1041. 157. Allen MJ, Thompson WD, Stuart RC, Gill PT, Walton EW, Karczenski K, Hoffman J & Peel ALG (1994) Management of nonpalpable breast lesions detected mammographically. Br J Surg 81: 543-545. 158. Homer MJ, Smith TJ & Safaii H (1992) Prebiopsy needle localization: methods, problems, and expected results. Radiol Clin North Am 30: 139-153. 159. Graham RA, Homer MJ, Sigler CJ, Safaii H, Schmid CH, Marchant DJ & Smith TJ (1994) The efficacy of specimen radiography in evaluating the surgical margins of impalpable breast carcinoma. AJR 162: 33-36. 160. Lee CH & Carter D (1995) Detecting residual tumor after excisional biopsy of impalpable breast carcinoma: efficacy of comparing preoperative mammograms with radiographs of the biopsy specimen. AJR 164: 81-86. 161. .Chadwick DR & Shorthouse AJ (1997) Wire-directed localization biopsy of the breast: an audit of results and analysis of factors influencing therapeutic value in the treatment of breast cancer. Eur J Surg Oncol 23: 128-133. 162. Acosta JA, Greenlee JA, Gubler KD, Goepfert CJ & Ragland JJ (1995) Surgical margins after needle-localization breast biopsy. Am J Surg 170: 643-646. 163. Stone MD & Cady B (1990) Techniques of lumpectomy and axillary dissection Surg Clin North Am 70: 885-899. 164. Jacobson JA, Danforth DN, Cowan KH, D'Angelo T, Steinberg SM, Pierce L, Lipp- man ME, Lichter AS, Glatstein E & Okunieff P (1995) Ten-year results of a comparison of conservation with mastectomy in the treatment of stage I and II breast cancer. N Eng J Med 332:907-911. 165. Osteen RT, Steele GD, Menck HR & Winchester DP (1992) Regional differences in surgical management of breast cancer. CA Cancer J Clin 42: 39-43. 166. Nylamo E, Hurme T & Lahti A (1998) Breast conservation or mastectomy in breast cancer: factors affecting the choice of operation. Ann Chir Gynaecol 87: 26-28. 167. Kotwall C, Covington D, Churchill P, Brinker C, Weintritt D & Maxwell JG (1998) Breast conservation surgery for breast cancer at a regional medical center. Am J Surg 176: 510-513. 168. Dolan JT & Granchi TS (1998) Low rate of breast conservation surgery in a large urban hospital serving the medically indigent. Am J Surg 176: 520-523. 169. Eberlein TJ (1994) Current management of carcinoma of the breast. Ann Surg 220: 121-136. 170. Tarbox BB, Rockwood JK & Abernathy CM (1992) Are modified radical mastectomies done for T1 breast cancers because of surgeons advice or patients choice? Am J Surg 164: 417-422. 66

171. Mann BA, Samet JM, Hunt WC, Key CR, Goodwin JM & Goodwin JS (1988) Changing treatment of breast cancer in New Mexico from 1969 through 1985. JAMA 259: 3413-3417. 172. Lazovich D, White E, Thomas D & Moe RE (1991) Underutilization of breast-conserving surgery and radiation therapy among women with stage I or II breast cancer. JAMA 266: 3433-3438. 173. Kotwall CA, Covington DL, Rutledge R, Churchill MP & Meyer AA (1996) Patient, hospital and surgeon factors associated with breast conservation surgery. Ann Surg 224: 419-429. 174. Blair SL, O'Shea KE & Orr RK (1997) Surgeon variability in treating nonpalpable breast cancer: surgical oncology as a value-added specialty. Ann Surg Oncol 5: 28- 32. 175. Clark J, Rosenman J, Cance W, Halle J & Graham M (1998) Extending the indications for breast-conserving treatment to patients with locally advanced breast cancer. IntJRadiatOncolBiolPhys42:345-350. 176. Hunt KK, Ames FC, Singletary SE, Buzdar AU & Hortobagyi GN (1996) Locally advanced noninflammatory cancer. Surg Clin North Am 76: 393-410. 177. The Steering Committee on Clinical Practice Guidelines for the Care and Treatment of Breast Cancer (1998) Mastectomy or lumpectomy? The choice of operation for clinical stages I and II breast cancer. Can Med Assoc J 158 Suppl 3: S15-21. 178. Ellis LM & Bland KI (1990) Techniques for obtaining the diagnosis of malignant breast lesions. Surg Clin North Am 70: 815-830. 179. Suomen Rintasyöpäryhmä ry (1997) Rintasyövän hoitosuositus. Duodecim 113:2341-2348. 180. Kearney TJ & Morrow M (1995) Effect of reexcision on the success of breast conserving surgery. Ann Surg Oncol 2: 303-307. 181. Papa MZ, Klein E, Davidson B, Karni T, Sperber F, Koller M, Ravid & Ari GB (1996) The effect of anesthesia type on needle localization breast biopsy: another point of view. Am J Surg 171: 242-243. 182. Margolese R, Poisson R, Shibata H, Pilch Y, Lerner H & Fisher B (1987) The technique of segmental mastectomy (lumpectomy) and axillary dissection: an syllabus from National Surgical Adjuvant Breast Project workshops. Surgery 102: 828-834. 183. Cady B (1995) How to perform breast biopsies. Surg Oncol Clin North Am 4: 47-66. 184. Ersek RA & Denton DR (1986) Breast biopsy technique: a plea for cosmesis. South Med J 79: 167-170. 185. Triedman SA, Osteen R & Harris JR (1990) Factors influencing cosmetic outcome of conservative surgery and radiotherapy for breast cancer. Surg Clin North Am 70: 901-916. 186. Morrow M (1995) Commentary. J Surg Oncol 58: 231-232. 187. Holmberg L (1995) Role of surgery in invasive cancer. Acta Oncol 34: 625-628. 188. Pandelidis SM, Peters KL, Walusimbi MS, Casady RL, Laux SV, Cavanaugh SH & Bauer TL (1997) The role of axillary dissection in mammographically detected carcinoma. J Am Coll Surg 184: 341-345. 189. Snider H, Dowlatshahi K, Fan M, Bridger WM, Rayudu G & Oleske D (1998) Sen- tinel node biopsy in the staging of breast cancer. Am J Surg 176: 305-310. 190. Veronesi U (1999) The sentinel node and breast cancer. Br J Surg 86: 1-2. 67

191. Fisher ER, Sass R, Fisher B, Gregorio R, Brown R, Wickerham L & Collaborating NSABP Investigators (1986) Pathologic findings from the National Surgical Adju- vant Breast Project (Protocol 6). II. Relation of local breast recurrence to multicentricity. Cancer 57: 1717-1724. 192. Schnitt SJ, Conolly JL, Khettry U, Mazoujian G, Brenner M, Silver B, Rect A, Beadle G & Harris JR (1987) Pathologic findings on re-excision of the primary site in breast cancer patients considered for treatment by primary radiation therapy. Can- cer 59: 675-681. 193. Frazier TG, Wong RWY & Rose D (1989) Implications of accurate pathologic margins in the treatment of primary breast cancer. Arch Surg 124: 37-38. 194. Solin LJ, Fowble B, Martz K, Pajak TF & Goodman RL (1986) Results of re-excisional biopsy of the primary tumor in preparation for definitive irradiation of patients with early stage breast cancer. Int J Radiat Oncol Biol Phys 12: 721-725. 195. Smitt MC, Nowela KW, Zdeblick MJ, Jeffrey S, Carlson RW, Stockdale FE & Goffi- net DR (1995) The importance of the lumpectomy surgical margin status in long term results of breast conservation. Cancer 76: 259-267. 196. Neff PT, Bear HD, Pierce CV, Grimes MM, Fleming MD, Neifeld JP, Arthus D, Horsley III JS, Lawrence Jr W & Kornstein MJ (1996) Long-term results of breast conservation therapy for breast cancer. Ann Surg 223:709-717. 197. Ryoo MC, Kagan AR, Wollin M, Tome MA, Tedeschi MA, Rao AR, Hintz BL, Kuruvilla AM, Nussbaum H, Streeter Jr OE, Jabola BR & Miller MJ (1989) Prog- nostic factors for recurrence and cosmesis in 393 patients after radiation therapy for early mammary carcinoma. Radiology 172: 555-559. 198. Spivack B, Khanna MM, Tafra L, Juillard G & Giuliano AE (1994) Margin status and local recurrence after breast-conserving surgery. Arch Surg 129: 952-957. 199. Harris JR, Connolly JL, Schnitt SJ, Cady B, Love S, Osteen RT, Patterson B, Shirley R, Hellman S, Cohen RB & Silen W (1985) The use of pathologic features in selecting the extent of surgical resection necessary for breast cancer patients treated by primary radiation therapy. Ann Surg 201: 164-169. 200. Solin LJ, Fowble BL, Schultz DJ & Goodman RL (1991) The significance of the pathology margins of the tumor excision on the outcome of patients treated with definitive irradiation for early stage breast cancer. Int J Radiat Oncol Biol Phys 21: 279-287. 201. Schmidt-Ullrich R, Wazer D, Tercilla O, Safai H, Marchant DJ, Smith TJ, Homer MA & Robert NJ (1989) Tumor margin assessment as a guide to optimal conservation surgery and irradiation in early stage breast carcinoma. Int J Radiat Oncol Biol Phys 17: 733-738. 202. Silverstein MJ, Lagios MD, Groshen S, Waisman JR, Lewinsky BS, Martino S, Gamagami P & Colburn WJ (1999) The influence of margin width on local control of ductal carcinoma in situ of the breast. N Engl J Med 340: 1455-1461. 203. Zafrani B, Vielh P, Fourquet A, Mosseri V, Durand JC, Salmon RJ & Vilcoq JR (1989) Conservative treatment of early breast cancer: prognostic value of ductal in situ component and other pathological variables on local control and survival. Eur J Cancer Clin Oncol 25: 1645-1650. 68

204. Fourquet A, Campana F, Zafrani B, Mosseri V, Vielh P, Durand J-C & Vilcoq JR (1989) Prognostic factors of breast recurrence in the conservative management of early breast cancer: a 25-year follow-up. Int J Radiat Oncol Biol Phys 17: 719-725. 205. Anscher MS, Jones P, Prosnitz LR, Blackstock W, Hebert M, Reddick R, Tucker A, Dodge R, Leight Jr G, Igiehart JD & Rosenman J (1993) Local failure and margin status in early-stage breast carcinoma treated with conservation surgery and radiation therapy. Ann Surg 218: 22-28. 206. McCormick B, Kinne D, Petrek J, Osborne M,Cox L, Shank B, Hellman S, Yahalom J & Rosen PP (1987) Limited resection for breast cancer: a study of inked specimen margins before radiotherapy. Int J Radiat Oncol Biol Phys 13: 1667-1671. 207. Ngai JH, Zelles GW, Rumore GJ, Sawicki JE & Godfrey RS (1991) Breast biopsy techniques and adequacy of margins. Arch Surg 126: 1343-1347. 208. Ghossein NA, Alpert S, Barba J, Pressman P, Stacey P, Lorenz E, Shulman M & Sadarangani GJ (1992) Importance of adequate surgical excision prior to radiotherapy in the local control of breast cancer in patients treated conservatively. Arch Surg 127: 411-415. 209. Bartelink H, Borger JH, van Dongen JA & Peterse JL (1988) The impact of tumor size and histology on local control after breast-conserving therapy. Radiother Oncol 11: 297-303. 210. Brenner MJ, Schnitt SJ, Connolly JL, Mazoujian G, Khettry U, Silver B, Beadle G, Recht A, Cohen RB & Harris JR (1985) The use of re-excision in primary radiation therapy for stage I and II breast carcinoma (abstr.). Int J Radiat Oncol Biol Phys 11 (Suppl. 1):186. 211. Beck NE, Bradburn MJ, Vincenti AC & Rainsbury RM (1998) Detection of residual disease following breast-conserving surgery. Br J Surg 85: 1273-1276. 212. Kollias J, Gill PG, Beamond B, Rossi H, Langlois S & Vernon-Roberts E (1998) Cli- nical and radiological predictors of complete excision in breast-conserving surgery for primary breast cancer. Aust N Z J Surg 68: 702-706. 213. Beron PJ, Horwitz EM, Martinez AA, Wimbish KJ, Levine AJ, Gustafson G, Chen PY, Ingold JA & Vicini FA (1996) Pathologic and mammographic findings predicting the adequacy of tumor excision before breast-conserving therapy. AJR 167: 1409-1414. 214. Gwin JL, Eisenberg BL, Hoffman JP, Ottery FD, Boraas M & Solin LJ (1993) Inci- dence of gross and microscopic carcinoma in specimens from patients with breast cancer after re-excision lumpectomy. Ann Surg 729-734. 215. Jardines L, Fowble B, Schultz D, Mackie J, Buzby G, Torosian M, Daly J, Weiss M, Orel S & Rosato E (1995) Factors associated with a positive reexcision after excisional biopsy for invasive breast cancer. Surgery 118: 803-809. 216. Schnitt SJ, Connolly JL, Harris JR, Hellman S & Cohen RB (1984) Pathologic predictors of early local recurrence in stage I and II breast cancer treated by primary radiation therapy. Cancer 53: 1049-1057. 217. Carter D (1986) Margins of "lumpectomy" for breast cancer. Hum Pathol 17: 330- 333. 218. Macmillan RD, Purushotham AD & George WD (1996) Local recurrence after breast-conserving surgery for breast cancer. Br J Surg 83: 149-155. 69

219. Schwartz GF, Rosenberg AL, Danoff BF, Mansfield CM & Feig SA (1984) Lumpec- tomy and level I axillary dissection prior to irradiation for "operable" breast cancer. Ann Surg 200: 554-559. 220. Lichter AS (1983) Current status of primary breast cancer treatment with local excision plus radiation. Curr Con Oncol 5: 17-23. 221. Heimann R, Powers C, Halpem HJ, Michel AG, Ewing CA, Wyman B, Recant W & Weichselbaum RR (1996) Breast preservation in stage I and II carcinoma of the breast. The University of Chicago experience. Cancer 78: 1722-1730. 222. Gazet J-C (1996) Future prospects in limited surgery for early breast cancer. Semin Surg Oncol 12: 39-45. 223. Blichert-Toft M, Smola MG, Cataliotti L & O'Higgins NO (1997) Principles and guidelines for surgeons-management of symptomatic breast cancer. Eur J Surg Oncol 101-109. 224. Wazer DE, DiPetrillo T, Schmidt-Ullrich R, Weld L, Smith TJ, Marchant DJ & Robert NJ (1992) Factors influencing cosmetic outcome and complication risk after conservative surgery and radiotherapy for early-stage breast carcinoma. J Clin Oncol 10: 356-363. 225. Early Breast Cancer Trialists' Collaborative Group (1995) Effects of radiotherapy and surgery in early breast cancer. An overview of the randomized trials. N Engl J Med 333: 1444-1455. 226. Ikeda T, Enomoto K, Wada K, Takeshima K, Yoneyama K, Furukawa J, Watanabe Y, Mukai M & Kitajima M (1997) Frozen-section-guided breast-conserving surgery: implications of diagnosis by frozen section as a guide to determining the extent of resection. Surg Today 27: 207-212. 227. Gal R (1988) Scrape cytology assessment of margins of lumpectomy specimens in breast cancer. Acta Cytol 32: 838-839. 228. Holland PA, Gandhi A, Knox WF, Wilson M, Baildam AD & Bundred NJ (1998) The importance of complete excision in the prevention of local recurrence of ductal carcinoma in situ. Br J Cancer 77: 110-114. 229. Cox CE, Ku NN, Reintgen DS, Greenberg HM, Nicosia SV & Wangensteen S (1991) Touch preparation cytology of breast lumpectomy margins with histologic correlation. Arch Surg 126: 490-493. 230. Cox CE, Pendas S, Ku NN, Reintgen DS, Greenberg HS & Nicosia SV (1998) Local recurrence of breast cancer after cytological evaluation of lumpectomy. Am Surg 64: 533-537. 231. Noguchi M, Minami M, Earashi M (1995) Intraoperative histologic assessment of surgical margins and lymph node metastasis in breast-conserving surgery. J Surg Oncol 60: 185-190. 232. Association of Directors of Anatomic and Surgical Pathology (1993) Immediate management of mammographically detected breast lesions. Hum Pathol 24: 689- 690. 233. Senofsky GM, Davies RJ, Olson L, Skully P & Olshen R (1990) The predictive value of needle localization mammographically assisted biopsy of the breast. Surg Gynecol Obstet 171:361-365. 70

234. Campbell ID, Theaker JM, Royle GT, Coddington R, Carpenter R, Herbert A, Moore I, Rubin C, Taylor I & Guyer PB (1991) Impact of an extensive in situ component on the presence of residual disease in screen detected breast cancer. J R Soc Med 84: 652-656. 235. Marrujo G, Jolly PC & Hall MH (1986) Nonpalpable breast cancer: needle-localized biopsy for diagnosis and considerations for treatment. Am J Surg 151: 599-602. 236. Tinnemans JGM, Wobbes T, van der Sluis RF, Lubbers E-JC & de Boer HHM (1986) Multicentricity in nonpalpable breast carcinoma and its implications for treatment. Am J Surg 151: 334-338. 237. Rosenberg AL, Schwartz GF, Feig SA & Patchefsky AS (1987) Clinically occult breast lesions: localization and significance. Radiology 162: 167-170. 238. Schwartz GF, Patchefsky AS, Feig SA, Shaber GS & Schwartz AB (1980) Clinically occult breast cancer. Multicentricity and implications for treatment. Ann Surg 191: 8-12. 239. Schwartz GF, Patchefsky AS, Feig SA, Shaber GS & Schwartz AB (1980) Multi- centricity of nonpalpable breast cancer. Cancer 45: 2913-2916. 240. Holland R, Veling SHJ, Mravunac M and Hendriks JHCL (1985) Histologic multifocality of Tis, T1-2 breast carcinomas: implications for clinical trials of breast-conserving surgery. Cancer 56: 979-990. 241. Morimoto T, Okazaki K, Komaki K, Sasa M, Mori T, Tsuzuki H, Kamamura Y, Miki H & Monden Y (1993) Cancerous residue in breast-conserving surgery. J Surg Oncol 52: 71-76. 242. Van Limbergen E, Rijnders A, van der Schueren E, Lerut T & Christiaens R (1989) Cosmetic evaluation of breast conserving treatment for mammary cancer. 2. A quantitative analysis of influence of radiation dose, fractionation schedules and surgical teatment techniques on cosmetic results. Radiother Oncol 16: 253-267. 243. Taylor ME, Perez CA, Halverson KJ, Kuske RR, Philpott GW, Garcia DM, Morti- mer JE, Myerson RJ, Radford D & Rush C (1995) Factors influencing cosmetic results after conservation therapy for breast cancer. Int J Radiat Oncol Biol Phys 31: 753-764. 244. Olivotto IA, Rose MA, Osteen RT, Love S, Cady B, Silver B, Recht A & Harris JR (1989) Late cosmetic outcome after conservative surgery and radiotherapy: analysis of causes of cosmetic failure. In J Radiot Oncol Biol Phys 17: 747-753. 245. Harris JR, Levene MB, Svensson G & Hellman S (1979) Analysis of cosmetic results following primary radiation therapy for stages I and II carcinoma of the breast. Int J Radiat Oncol Biol Phys 5: 257-261. 246. Ray GR & Fish VJ (1983) Biopsy and definitive radiation therapy in stage I and II adenocarcinoma of the female breast: analysis of cosmesis and the role of electron beam supplementation. Int J Radiat Oncol Biol Phys 9: 813-818. 247. Matory WE, Jr, Wertheimer M, Fitzgerald TJ, Walton RL, Love S & Matory WE (1989) Aesthetic results following partial mastectomy and radiation therapy. Plast Reconstr Surg 85: 739-746. 248. Steeves RA, Phromratanapongse P, Wolberg WH & Tormey DC (1989) Cosmesis and local control after irradiation in women treated conservatively for breast cancer. Arch Surg 149: 1369-1373. 71

249. De la Rochefordiere A, Abner AL, Silver B, Vicini F, Recht A & Harris JR (1992) Are cosmetic results following conservative surgery and radiation therapy for early breast cancer dependent on technique? Int J Radiat Oncol Biol Phys 23: 925-931. 250. Clarke D, Martinez A & Cox RS (1983) Analysis of cosmetic results and complications in patients with stage I and II breast cancer treated by biopsy and irradiation. Int J Radiat Oncol Biol Phys 9:1807-1813. 251. Haukipuro K (1991) Synthesis of collagen types I and III in reincised wounds in humans. Br J Surg 78: 708-712. 252. Curran D, van Dongen JP, Aaronson NK, Kiebert G, Fentiman IS, Mignolet F & Bar- telink H (1998) Quality of life of early-stage breast cancer patients treated with radical mastectomy or breast-conserving procedures: results of EORTC Trial 10801. Eur J Cancer 34: 307-314. 253. Pierquin B, Otmezguine Y & Lobo PA (1983) Conservative management of breast carcinoma. The Créteil experience. Acta Radiol Oncol 22: 101-107. 254. Mills JM, Schultz & Solin LJ (1997) Preservation of cosmesis with low complication risk after conservative surgery and radiotherapy for ductal carcinoma in situ of the breast. Int J Radiat Oncol Biol Phys 39: 637-641. 255. Gray JR, McCormick B, Cox L & Yahalom J (1991) Primary breast irradiation in large-breasted or heavy women: analysis of cosmetic outcome. Int J Radiat Oncol Phys 21: 347-354. 256. Hermanek P & Sobin LH (eds) (1992) "TNM Classification of Malignant Tumours" Fourth Edition, 2nd Revision. Berlin: Springer- Verlag. 257. Aaronson NK, Bartelink H, van Dongen JA & van Dam FSAM (1988) Evaluation of breast conserving therapy: clinical, methodological and psychosocial perspectives. Eur J Surg Oncol 14: 133-140. 258. Jackson VP (1990) Mammographically guided fine-needle aspiration biopsy for nonpalpable breast lesions. Curr Opin Radiol 2: 741-745. 259. Tsoukas LI & Fentiman IS (1990) A new method for evaluation of cosmetic outcome after conservative treatment of early breast cancer. Breast Cancer Res Treat 15: 185-190. 260. Pezner RD, Lipsett JA, Vora NL & Desai KR (1985) Limited usefulness of observer- based cosmesis scales employed to evaluate patients treated conservatively for breast cancer. Int J Radiat Oncol Biol Phys 11: 1117-1119. 261. Pezner RD, Patterson MP, Hill LR, Vora N, Desai KR, Archambeau JO & Lipsett JA (1985) Breast retraction assessment: an objective evaluation of cosmetic results of patients treated conservatively for breast cancer. Int J Radiat Oncol Biol Phys 11: 575-578. 262. Van Limbergen E, van der Schueren & Van Tongelen K (1989) Cosmetic evaluation of breast conserving treatment for mammary cancer. 1. Proposal of a quantitative scoring system. Radiother Oncol 16: 159-167. 263. Klijanienko J, Cote JF, Thibault F, Zafrani B, Meunier M, Clough K, Asselain B & Vielh P (1998) Ultrasound-guided fine-needle aspiration cytology of nonpalpable breast lesions: Institut Curie's experience with 198 histologically correlated cases. Cancer 84: 36-41. 72

264. Boerner S, Fornage BD, Singletary E & Sneige N (1999) Ultrasound-guided fine- needle aspiration (FNA) of nonpalpable breast lesions: a review of 1885 FNA cases using the National Cancer Institute-supported recommendations on the uniform approach to breast FNA. Cancer 87: 19-24. 265. Fornage BD, Sneige N, Faroux M & Andry E (1990) Sonographic appearance and ultrasound-guided fine-needle aspiration biopsy of breast carcinomas smaller than 1 cm3. J Ultrasound Med 9:559-568. 266. Sickles EA (1991) Periodic mammographic follow-up of probably benign breast lesions: results in 3184 consecutive cases. Radiology 179: 463-468. 267. Wilkinson EJ, Schuettke CM, Ferrier CM, Franzini DA & Bland KI (1989) Fine needle aspiration of the breast masses: an analysis of 276 aspirates. Acta Cytol 33: 613-619. 268. EisenbergAJ, Hajdu SI, Wilhelmus J, Melamed MR & Kinne D (1986) Preoperative aspiration cytology of breast tumors. Acta Cytol 2: 135-146. 269. Wilkinson EJ & BLand KI (1990) Techniques and results of aspiration cytology for diagnosis of benign and malignant diseases of the breast. Surg Clin North Am 70: 801-813. 270. Wilkinson S, Anderson TJ, Rifkind E, Chetty U & Forrest AP (1989) Fibroadenoma of the breast: a follow-up of conservative management. Br J Surg 76: 390-391. 271. Grillo M, Lehmann-Willenbrock E & Gent H-J (1990) Ann Chir Gynaecol 79: 6-9. 272. Coburn MC & Bland KI (1995) Surgery for early and minimally invasive breast cancer. Curr Opin Oncol 7: 506-510. 273. Aitken RJ, Going JJ & Chetty U (1990) Assessment of surgical excision during breast conservation surgery by intraoperative two-dimensional specimen radiology. Br J Surg 77: 322-323. 274. Cox CE, Reintgen DS, Nicosia SV, Ku NN, Baekey P & Carey LC (1995) Analysis of residual cancer after diagnostic breast biopsy: an argument for fine-needle aspiration cytology. Ann Surg Oncol 2: 201-206. 275. Sarin R, Dinshaw KA, Shrivastava SK, Sharma V & Deore SM (1993) Therapeutic factors influencing the cosmetic outcome and late complications in the conservative management of early breast cancer. Int J Radiat Oncol Biol Phys 27: 285-292. 276. Krishnan L, Jewell WR, Mansfield CM, Reddy EK, Thomas JH, Krishnan EC & Gerald K (1988) Cosmetic results in early breast cancer treated with lumpectomy, peri-operative interstitial irradiation and external beam radiation. Int J Radiat Oncol Biol Phys 14: 205-211. 277. Witte MB & Barbul A (1997) General principles of wound healing. Surg Clin North Am 77: 509-528. 278. Gould EW & Robinson PG (1992) The pathologist's examination of the "lumpectomy"- the pathologists' view of surgical margins. Semin Surg Oncol 8: 129-135. 279. Holmberg L, Zarén E, Adami H-O, Bergström R & Burns T (1988) The patient's appraisal of the cosmetic result of segmental mastectomy in benign and malignant breast disease. Ann Surg 207: 189-194.