ARTICLE IN PRESS

The Breast (2005) 14, 523–526

THE

BREAST

KEYWORDBreast contherapy;LumpectomMagnetic rimaging

0960-9776/$ - sdoi:10.1016/j.b

�Tel.: +1 215E-mail addr

www.elsevier.com/locate/breast

ORIGINAL ARTICLE

Limiting breast surgery to the proper minimum

Monica Morrow�

Department of Surgical Oncology, Fox Chase Cancer Center, 333 Cottman Avenue, Philadelphia,PA 19111-2497, USA

Sserving

y;esonance

ee front matter & 2005reast.2005.08.008

728 3096; fax: +1 215 21ess: monica.morrow@fc

Summary Clinical experience has resulted in the identification of a relativelysmall number of absolute contraindications to breast-conserving therapy (BCT).These contraindications are readily identified by a clinical evaluation and diagnosticmammography. Local failure rates of less than 10% at 10 years support the idea thatpatients can be reliably selected for BCT with standard clinical modalities. Theavailability of magnetic resonance (MR) imaging has raised questions about its role inpatient selection for BCT. MR imaging identifies additional cancer in 10–54% ofpatients with apparently localized disease, resulting in mastectomies that would nototherwise have been done. Clinical experience suggests that the majority of thisdisease is controlled by radiotherapy. Studies demonstrating clinical benefit in termsof decreased rates of local recurrence or fewer surgeries are needed before MR isused for routine selection of patients for BCT.& 2005 Elsevier Ltd. All rights reserved.

Introduction

The appropriate surgical therapy for breast canceris dependent upon the reliable identification ofpatients for whom mastectomy is medically neces-sary, the resection of an adequate amount of breasttissue to maintain local control and preservecosmesis in patients undergoing breast-conservingtherapy (BCT), and an assessment of patientpreferences.

Clinical experience has identified a limitednumber of absolute contraindications to the useof BCT. These include first- or second-trimester

Elsevier Ltd. All rights reserv

4 4035.cc.edu.

pregnancy, a history of prior irradiation to thebreast field, multicentric carcinoma, and theinability to obtain negative surgical margins.1 Thepresence of scleroderma or active lupus erythema-tosus is considered a relative contraindication. Howcommonly contraindications to BCT occur has notbeen well studied.

Review

Morrow et al.2 reported that in 456 consecutivepatients evaluated by a multidisciplinary team at areferral institution, 21% had contraindications toBCT. The likelihood of contraindications varied bystage, with only 10% of stage I patients having

ed.

ARTICLE IN PRESS

M. Morrow524

contraindications compared with 28% of stage IIpatients and 33% of those with ductal carcinoma insitu (DCIS; Po0:01). In a population-based study of1884 women with DCIS and Stage I and II breastcancer treated between December 2001 andJanuary 2003, Katz et al.3 reported a 30% mas-tectomy rate, with 11.5% of patients specificallynoting contraindications to BCT.

There is evidence that patients with contra-indications to BCT can be reliably identified on thebases of a history, physical examination, anddiagnostic mammography. Morrow et al.4 reportedthat in 216 patients thought to be candidates forBCT based on this type of evaluation, the procedurewas successfully carried out in 210 (97.3%). In spiteof this high rate of success in selecting patients forBCT with conventional, low-cost imaging, there isgreat interest in using newer imaging modalitiessuch as whole breast ultrasound (US) and magneticresonance (MR) to select patients for BCT. Clinicaland imaging evaluation is also used to define theextent of the breast tissue to be resected inpatients undergoing BCT. There is great variabilityin the likelihood of achieving negative margins witha single resection, no consensus on what constitu-tes the ideal amount of normal breast tissue toexcise, and the incidence of margin positivity hasbeen reported to vary with the method of tumordetection, histologic tumor type, patient age,tumor size, surgeon experience, and method ofpathologic examination.5,6 This combination offactors makes it difficult to determine the influenceof an imaging modality on margin status outside ofa large prospective trial, so this paper will focus onthe role of imaging in patient selection for BCT.

There have been a small number of studies on theimpact of whole breast US on eligibility for BCT.Berg and Gilbreath 7 reported that US depictedunsuspected multifocal carcinoma in three out of

Table 1 Detection of multifocal/centric cancer by magn

Reference Year

Harms et al.10 1993Boetes et al.11 1995Mumtaz et al.12 1997Fischer et al.13 1999Drew et al.14 1999Esserman et al.15 1999Liberman et al.16 2003Furman et al.17 2003Bedrosian et al.18 2003Schnall et al.19 2004Berg et al.20 2004

20 (15%) patients, and Moon et al.8 reported that14% of 201 patients had occult multifocal ormulticentric disease identified with US. In contrast,Golshan et al.9 found that while 75 out of 426 (18%)patients had additional abnormalities identified byUS, only 12 (2.8%) were malignant. The mostcommon change in therapy on the basis of USfindings was mastectomy. Although patients withUS-detected abnormalities were younger and weremore likely to have clinically detected cancers ofhigher grade than those without US abnormalities,clinical and tumor factors distinguishing betweenthose with benign and malignant US findings couldnot be identified.9

Although US is a low-cost modality that is welltolerated by patients, much greater attention hasbeen focused on the use of MR in breast cancertreatment selection. The results of multiple studiesdemonstrating that MR identifies additional foci ofcancer in 11–54% of patients thought to haveunicentric disease after clinical and mammographicevaluation are summarized in Table 1. 10–20 In amulti-institutional study involving 14 sites and 426cancers, Schnall et al.19 defined MR-detected‘‘incidental lesions’’ as those at a distance ofgreater than 2 cm from the known primary cancer.Incidental lesions were found in 103 patients (24%)and 78 were biopsied, resulting in the identificationof 56 additional cancers. In this study, MR poten-tially changed treatment for 10% of the studypopulation, leading the authors to conclude that itshould be a routine part of the evaluation ofpatients prior to the selection of local therapy. Twostudies have directly assessed the impact of MR onsurgical treatment. Bedrosian et al.18 evaluated267 patients with T1–T4 tumors with MR and foundabnormalities in 26%. The MR findings resulted inmastectomy in 44 patients (16.5%), a wider excisionin 11 (4%), and a separate excisional biopsy in the

etic resonance imaging.

n Additional cancer (%)

29 5461 1592 11

336 16178 2358 1070 1976 13

267 15426 2496 30

ARTICLE IN PRESS

Limiting breast surgery to the proper minimum 525

remainder. Of note, malignancy was verified histo-logically in only 71% of cases. Berg et al.20 reportedon the clinical consequences of MR in 121 cancerpatients. Findings on MR resulted in unnecessarymastectomies in 12% and the need for follow-upimaging studies in half of the group. In the subset of96 women felt to be candidates for BCTon the basisof clinical and mammographic evaluation, 29 (30%)were found to have additional cancer, butthe extent of disease was overestimated in 20of the 29.

The high incidence of additional areas of cancerseen only with MR imaging has led many radiologiststo conclude that MR is an essential part of theevaluation of patients being considered for BCT.However, the identification of additional cancer ina quarter to a third of patients is difficult toreconcile with the clinical observation that 90–95%of patients selected for BCT on the basis of clinicaland mammographic evaluation are free of localfailure at 10 years in modern series.5,21,22 Theargument that MR findings are an indication formastectomy is reminiscent of the debate in the1970s that all breast cancer patients requiredtreatment by mastectomy because pathologicstudies using serial subgross sectioning demon-strated microscopic carcinoma beyond the primarytumor in 21–63% of patients with apparentlyunicentric cancer.23–30 These studies are summar-ized in Table 2. The work of Holland et al.30 hasshown that the majority of these tumor foci arefound within 4 cm of the primary tumor and areunlikely to be completely removed with a limitedpartial mastectomy or lumpectomy. This pattern ofdisease distribution is consistent with what isdescribed in the majority of studies onMR.13,14,16,20 There are two potential ways toreconcile the seemingly contradictory findings ofthe large amounts of cancer identified by MR andthe high rates of local control after BCT in patientsselected using mammography and clinical evalua-tion. First, a significant amount of the MR-detected

Table 2 Pathologic studies of multifocality/multicentric

Reference No. of cases Population

Qualheim and Gall23 157 Not statedRosen et al.24 203 Invasive carcinLagios25 85 Not statedEgan26 118 Not statedSchwartz et al.27 43 Non-palpable cVaidya et al.28 30 Invasive carcinAnastassiades et al.29 366 Invasive o7 cmHolland et al.30 282 Clinically unice

disease may be removed surgically because it isdetected intraoperatively or results in positivemargins of excision. If this is the case, determina-tion of whether or not MR decreases the number ofsurgical procedures needed to achieve negativemargins becomes critical in determining its role intreatment selection. Alternatively, most of the MR-detected disease may be of low enough volumethat it can be controlled with radiotherapy. Thishypothesis is consistent with early studies on BCTusing radiotherapy alone for T1 tumors, whichresulted in local control rates equal to those seenwith excision and irradiation.31 These explanationsare not mutually exclusive and it is mostly likelythat both apply. With MR we now have technologycapable of identifying tumor foci that couldpreviously be found only on pathologic examina-tion, but the clinical implications of their identifi-cation are uncertain. To date, no studies addressingthe impact of MR selection for BCT on outcomessuch as local recurrence or the number of surgicalprocedures needed to obtain negative margins havebeen carried out. Such studies are essential prior tothe routine adoption of MR, since the use of thetechnology increases the rate of mastectomy,results in a significant number of benign biopsies,and adds costs.

Conclusions

Patients can be reliably selected for BCT usingclinical evaluation and mammography. Over thepast 20 years, rates of local recurrence after BCThave decreased with improved mammographic,pathologic, and surgical evaluation and the in-creased use of adjuvant systemic therapy. Demon-stration of a benefit of MR in the general populationof patients undergoing BCT will require very largeand expensive studies due to the low rates of localfailure that are currently seen. A better utilization

ity.

% Multifocal/centric

54oma 33

2160

ancer 44oma 63, noninvasive 49ntric invasive cancer o5 cm 63

ARTICLE IN PRESS

M. Morrow526

of resources would be to evaluate the benefit of MRin areas where the appropriate extent of surgeryremains problematic. These include identificationof patients with positive or close margins whowould benefit from re-excision and an improveddefinition of the volume of tissue that must beresected for BCT in patients who have responded toneoadjuvant therapy. In addition, MR may allow theidentification of patients who are appropriatecandidates for partial breast irradiation or thosewho do not require radiotherapy as part of breast-conserving therapy.

References

1. Morrow M, Strom EA, Bassett LW, et al. Standard for breastconservation therapy in the management of invasive breastcarcinoma. CA Cancer J Clin 2002;52:277–300.

2. Morrow M, Bucci C, Rademaker A. Medical contraindicationsare not a major factor in the underutilization of breastconserving therapy. J Am Coll Surg 1998;186:269–74.

3. Katz SJ, Lantz PM, Janz NK, et al. Patient involvement insurgery treatment decisions for breast cancer. J Clin Oncol2005;23:5526–33.

4. Morrow M, Schmidt R, Hassett C. Patient selection for breastconservation therapy with magnification mammography.Surgery 1995;118:621–6.

5. Park CC, Mitsumori M, Nixon A, et al. Outcome at 8 yearsafter breast-conserving surgery and radiation therapy forinvasive breast cancer: influence of margin status andsystemic therapy on local recurrence. J Clin Oncol 2000;18:1668–75.

6. Peterson ME, Schultz DJ, Reynolds C, Solin LJ. Outcomes inbreast cancer patients relative to margin status aftertreatment with breast-conserving surgery and radiationtherapy: the University of Pennsylvania experience. Int JRadiat Oncol Biol Phys 1999;43:1029–35.

7. Berg WA, Gilbreath PL. Multicentric and multifocal cancer:whole-breast US in preoperative evaluation. Radiology 2000;214:59–66.

8. Moon WK, Noh DY, Im JG. Multifocal, multicentric, andcontralateral breast cancers: bilateral whole-breast US inthe preoperative evaluation of patients. Radiology 2002;224:569–76.

9. Golshan M, Fung BB, Wolfman J, Rademaker A, Morrow M.The effect of ipsilateral whole breast ultrasonography on thesurgical management of breast carcinoma. Am J Surg 2003;186:391–6.

10. Harms SE, Flamig DP, Hesley KL, et al. MR imaging of thebreast with rotating delivery of excitation off resonance:clinical experience with pathologic correlation. Radiology1993;187:493–501.

11. Boetes C, Mus RD, Holland R, et al. Breast tumors:comparative accuracy of MR imaging relative to mammo-graphy and US for demonstrating extent. Radiology 1995;197:743–7.

12. Mumtaz H, Hall-Craggs MA, Davidson T, et al. Staging ofsymptomatic primary breast cancer with MR imaging. AJRAm J Roentgenol 1997;169:417–24.

13. Fischer U, Kopka L, Grabbe E. Breast carcinoma: effect ofpreoperative contrast-enhanced MR imaging on the ther-apeutic approach. Radiology 1999;213:881–8.

14. Drew PJ, Turnbull LW, Chatterjee S, et al. Prospectivecomparison of standard triple assessment anddynamic magnetic resonance imaging of the breast for theevaluation of symptomatic breast lesions. Ann Surg 1999;230:680–5.

15. Esserman L, Hylton N, Yassa L, Barclay J, Frankel S, SicklesE. Utility of magnetic resonance imaging in the managementof breast cancer: evidence for improved preoperativestaging. J Clin Oncol 1999;17:110–9.

16. Liberman L, Morris EA, Dershaw DD, Abramson AF, Tan LK.MR imaging of the ipsilateral breast in women withpercutaneously proven breast cancer. AJR Am J Roentgenol2003;180:901–10.

17. Furman B, Gardner MS, Romilly P, et al. Effect of 0.5 Teslamagnetic resonance imaging on the surgical management ofbreast cancer patients. Am J Surg 2003;186:344–7.

18. Bedrosian I, Mick R, Orel SG, et al. Changes in the surgicalmanagement of patients with breast carcinoma based onpreoperative magnetic resonance imaging. Cancer2003;98:468–73.

19. Schnall M, Blume J, Bluemke D, et al. MRI detection ofmultifocal breast carcinoma: report from the InternationalBreast MRI Consortium. J Clin Oncol 2004;22:4s (abstr. 504).

20. Berg WA, Gutierrez L, NessAiver MS, et al. Diagnosticaccuracy of mammography, clinical examination, US, andMR imaging in preoperative assessment of breast cancer.Radiology 2004;233:830–49.

21. Freedman G, Fowble B, Hanlon A, et al. Patients with earlystage invasive cancer with close or positive margins treatedwith conservative surgery and radiation have an increasedrisk of breast recurrence that is delayed by adjuvantsystemic therapy. Int J Radiat Oncol Biol Phys 1999;44:1005–15.

22. Dewar JA, Arriagada R, Benhamou S, et al. Local relapse andcontralateral tumor rates in patients with breast cancertreated with conservative surgery and radiotherapy (InstitutGustave Roussy 1970–1982). IGR Breast Cancer Group.Cancer 1995;76:2260–5.

23. Qualheim RE, Gall EA. Breast carcinoma with multiple sitesof origin. Cancer 1957;10:460–8.

24. Rosen PP, Fracchia AA, Urban JA, Schottenfeld D, RobbinsGF. Residual mammary carcinoma following simulatedpartial mastectomy. Cancer 1975;35:739–47.

25. Lagios MD. Multicentricity of breast carcinoma demon-strated by routine correlated serial subgross and radio-graphic examination. Cancer 1977;40:1726–34.

26. Egan RL. Multicentric breast carcinomas: clinical-radio-graphic-pathologic whole organ studies and 10-year survival.Cancer 1982;49:1123–30.

27. Schwartz GF, Patchesfsky AS, Feig SA, Shaber GS, SchwartzAB. Multicentricity of non-palpable breast cancer. Cancer1980;45:2913–6.

28. Vaidya JS, Vyas JJ, Chinoy RF, Merchant N, Sharma OP, MittraI. Multicentricity of breast cancer: whole-organ analysis andclinical implications. Br J Cancer 1996;74:820–4.

29. Anastassiades O, Iakovou E, Stavridou N, Gogas J, KaramerisA. Multicentricity in breast cancer. A study of 366 cases. AmJ Clin Pathol 1993;99:238–43.

30. Holland R, Veling SH, Mravunac M, Hendriks JH. Histologicmultifocality of Tis, T1-2 breast carcinomas. Implications forclinical trials of breast-conserving surgery. Cancer 1985;56:979–90.

31. Amalric R, Santamaria F, Robert F, et al. Radiation therapywith or without primary limited surgery for operable breastcancer: a 20-year experience at the Marseilles CancerInstitute. Cancer 1982;49:30–4.