The Breast xxx (2015) 1e6
Contents lists available at ScienceDirect
The Breast journal homepage: www.elsevier.com/brst
Original article
Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer Caroline Malhaire a, *, Delphine Hequet b, Marie-Christine Falcou c, nie Guillot b, Jean-Guillaume Feron b, Anne Tardivon a, Alexandre Leduey b, Euge c b b ronique Mosseri , Roman Rouzier , Benoit Couturaud , Fabien Reyal b Ve a b c
Department of Radiology, Institut Curie, 26 rue d’Ulm, 75005 Paris, France Department of Surgical Oncology, Institut Curie, 26 rue d’Ulm, 75005 Paris, France Department of Public Health, Institut Curie, 26 rue d’Ulm, 75005 Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 November 2014 Received in revised form 16 February 2015 Accepted 22 February 2015 Available online xxx
Purpose: The purpose of this study was to evaluate the outcome of breast conserving surgery comparing oncoplastic surgery (OS) and standard lumpectomy (SL) after preoperative bracketing wire localization of large neoplastic lesions. Methods: We retrospectively reviewed the medical records and the mammograms of patients operated on at the Institut Curie between May 2005 and September 2011 after bracketing wire localization under mammographic and/or sonographic guidance. Results: 113 patients underwent surgery for a pre-operative diagnosis of DCIS (n ¼ 80), micro-invasive carcinoma (n ¼ 9) or invasive carcinoma (n ¼ 24), by OS (n ¼ 73) or SL (n ¼ 40). In the OS group, radiological size (52 mm vs 39 mm, p < 0.001) and resection volumes (246 cc vs 88 cc, p < 0.00001) were significantly higher than in the SL group. Rates of clear histologic margins (60 vs 62%, NS), complete excision of microcalcifications (78% vs 72%, NS) and re-intervention rate (40% vs 42%, NS) were equivalent. The rate of local recurrence at 24 months was 3% [0e7.1] in patients with conservative treatment (n ¼ 3). With a median follow-up of 40 months, 5 local relapses (two with axillary metastatic involvement), two distant metastatic evolution, one contralateral breast cancer and one death unrelated to cancer occurred. Conclusion: Following bracketing wire localization, OS allowed the conserving management of significantly larger lesions with wider resection volumes, without significant increase in margin involvement or re-intervention rate, and equivalent rate of microcalcifications clearance compared to SL. © 2015 Elsevier Ltd. All rights reserved.
Introduction Breast-conservation surgery (BCS) allows equivalent survival outcome to that of mastectomy, although the rate of local recurrence is slightly higher [1,2]. BCS involves performing a resection wide enough to ensure margins clearance and local control, while maintaining acceptable cosmetic outcome. Despite this, poor cosmetic outcomes are still reported in up to 40% of patients after BCS [3]. Rather than an absolute tumour dimension, ratio of tumour-to-breast and tumour location are determinant keys for the indication of conservative surgery.
* Corresponding author. Tel.: þ33 1 44 32 42 81. E-mail address: [email protected] (C. Malhaire).
Recent years have seen the emergence of oncoplastic techniques combining plastic surgery procedures of the breast to standard lumpectomy. Oncoplastic surgery (OS) relies on advancement, rotation or transposition of a breast-flap [4], allowing for a resection of lesions that would otherwise not be amenable to standard lumpectomy, while also allowing for better cosmetic results [5]. Large calcifications clusters revealing breast cancer is a challenging situation regarding breast conservation. Extensive ductal carcinoma in situ (DCIS) is frequently growing in a radial fashion [6], as a contiguous disease in the same segment of the breast [7] and may thereby benefit from oncoplastic techniques, as it is not appropriate for standard lumpectomy (SL). Multiple wire localization techniques have been recommended to bracket large calcifications clusters before performing BCS in order to better document
http://dx.doi.org/10.1016/j.breast.2015.02.037 0960-9776/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
2
C. Malhaire et al. / The Breast xxx (2015) 1e6
lesion extension [8]. The aim of our study was to assess the outcome of BCS with wide local excision of breast carcinoma using bracketing wire localization (BWL), regarding complete excision of calcifications at post-operative mammogram, margin status, reoperation and completion mastectomy rates. Materials and methods Local ethical committee approved retrospective review of patients' records and imaging studies. From our radiological report data of May 2005 to September 2011, we retrospectively identified 188 consecutive BWL performed under mammographic and/or sonographic guidance for breast carcinoma. Exclusion criteria included benign or atypical lesions (n ¼ 42), neoadjuvant chemotherapy (n ¼ 16), wire localization performed for distinct lesions (n ¼ 17). A total of 113 cases with a preoperative needle biopsy diagnosis of DCIS and/or invasive carcinoma were finally included. All mammograms were retrospectively reviewed by a radiologist trained to breast imaging (CM, 10 years experience) without knowledge of post-operative results. Breast density according to the BI-RADS classification, type and distribution of calcifications and two-dimensional size of the calcifications cluster were recorded using BI-RADS lexicon [9]. The maximal axis size of the lesion measured on each view was retained for analysis. BWL was performed under mammographic or sonographic guidance for lesions considered to be too large to be accurately marked by single wire localization. Usually, a lesion of 3 cm or more was regarded as an indication for bracketing wire localization in our institution [10]. Five or 9 cm length wires (Medi-tech/Boston Scientific, Watertown, Mass and Cook, Bloomington, IA) were placed in an attempt to mark out the lesion and define as precisely as possible the upperlower and latero-medial extension. As much as possible, wire needles were inserted from the same side of the breast with the shortest approach to facilitate surgical dissection. Wire localization was performed using a compression paddle with an opening area (Mammomat 3000, Siemens, Erlangen Germany and Senograph Essential, GE Healthcare, Milwaukee, Wisconsin). Latero-medial and cranio-caudal mammograms were systematically performed after wire deployment to assess the position of wires relative to the calcifications cluster (Fig. 1AeC). A metallic dot was placed to help surgeons identify the nipple. After resection, a radiograph of the oriented specimen was performed to document successful removal of breast calcifications and to assess radiographic margins (Fig. 1D). After resection of large calcifications clusters, post-operative mammogram was recommended [11]. The largest dimension of the lesion was determined from pathology reports and volume was calculated using the specimen's length, height and width. Clear histological margins were defined as superior to 1 mm [12]. Date and type of secondary surgery and histological results were recorded, as well as follow-up medical records of local or distant relapse, and second cancers. OS was performed by either plastic surgeons, or breast surgeons with OS training. Statistical analysis Categorical variables were summarized as counts and percentages in each class, the quantitative variables as mean and standard deviation, median, minimum, maximum. The observed percentages were compared by the chi-square test or Fisher's exact test. Medians were compared using the nonparametric Kruskal Wallis test. The differences were considered significant if p is less than 0.05.
Results The median age was 58 years [range 38e63]. Four patients had high familial risk according to genetic counseling, without mutation identification in three of them. Tests results were still pending for the fourth patient. Median maximal dimension of calcifications clusters measured on preoperative mammograms was 47 mm overall [range, 15e115 mm], significantly larger in oncoplastic group (median 52 mm, range: 15e92) than in SL group (median 39 mm, range: 19e115, p < 0.01). In 7 patients, pre-operative mammograms were not available to assess calcifications cluster size. None of the radiographic features reported correlated with positive margins (Table 1). Pre-operative biopsy (Table 2) showed DCIS in 80, microinvasive DCIS in 9 and invasive carcinoma in 24. Of 113 patients, 73 underwent OS. Surgical techniques used during OS are summarized in Table 3 [13]. Comparative results of SL and OS are summarized in Table 4. The median resection volume (not available in 17 cases) was 172 cc3 overall [range 180e1980 cc3, mean 298 cc3, standard deviation 385]. significantly larger in the oncoplastic group with a median of 246 cc3 [range 42e1980 cc3] versus a median of 88.5 cc3 [range 18e448 cc3] in the SL group (p < 0.0001). Seventeen out of 80 lesions (21%) diagnosed as DCIS at percutaneous biopsy appeared to be invasive at final operative histologic analysis, while 6 of the 9 micro-invasive carcinomas were found to be invasive at surgical histologic analysis. Median size was 9 mm [range 2e120 mm] for invasive carcinomas and 35 mm [range 6e100 mm] for DCIS. Clear histologic margins were obtained in 61% with a median margin of 3.5 mm [range 1e20 mm, mean 4 mm]. In the case of positive margins, extent of margin involvement (available in n ¼ 40) was focal (15 mm) in 3 and multifocal in 12. Lesions resected with positive margins had a median mammographic size of 50 mm [range 19e115 mm], whereas those with negative margins had a median mammographic size of 46 mm [range 15e92 mm] (NS, p ¼ 0.15, Student t test). Margins status was not significantly associated with type of surgery (p ¼ 0.82, chi-square), as shown in Table 4. Fifty-seven of 75 patients with post-operative mammograms (76%) did not show any residual calcifications. Twenty patients underwent immediate local re-excision because of positive margins at histological analysis without post-operative mammogram. Forty-seven patients underwent local re-excision by BCS (n ¼ 15) or mastectomy (n ¼ 32) about half of which associated with immediate reconstruction. Overall, 31 patients had no post-operative radiotherapy. 26 had post-operative radiotherapy without local boost radiation and 56 patients underwent boost radiation following surgery, of whom 4 had close or positive margins, and 52 had final clear histologic margins (considering eventual local re-excision). The rate of local recurrence at 24 months was 3% [0e7.1] in patients with BCS. During a median of 40 months' follow-up assessment [1e91 months], two patients died from metastatic evolution, and one death was unrelated to cancer evolution. Local relapses, detailed below, occurred in 5 patients in the same quadrant as index cancer among the 81 patients who underwent BCS (6%), of whom two had axillary metastatic involvement. No relapse was recorded in the mastectomy group. One contralateral breast cancer was recorded. Patient 1. Local relapse occurred in the same lower-outer quadrant 26 months after the calcifications cluster had been bracketed and the patient operated on by quadrantectomy. Inferior margin was focally (5 mm) less than 1 mm for invasive and noninvasive carcinoma. She underwent whole breast radiation
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
C. Malhaire et al. / The Breast xxx (2015) 1e6
3
Fig. 1. 52-year-old woman with a preoperative diagnosis of IDC and DCIS from the US-guided microbiopsy of a hypoechoic nodule, associated with extensive calcifications at mammography. Microcalcification cluster extended 5 cm antero-posteriorly and 6 cm large on the cranio-caudal magnification view (A). Cranio-caudal view (B) and medio-lateral view (C) show BWL, one posterior wire indicating lateral and upper limit of the microcalcification cluster and another anterior wire indicating inner and lower limit of the microcalcification cluster. A J-plasty OS was performed allowing the resection of a 146 g specimen containing calcifications in its center at intraoperative radiograph (D). Wires are visible in the specimen, delineating lesion borders. No residual calcification of the large cluster was visible on cranio-caudal at postoperative mammogram. Histological analysis showed bifocal invasive ductal carcinoma of 18 and 5 mm (grade II, low mitotic index) associated with extensive DCIS. Minimal interval to lateral margins was 3 mm. Patient underwent chemotherapy, radiotherapy of the breast with local boost complement, radiation of the internal mammary, sus and sub clavian nodes and hormonotherapy (Aromasine®). She is free from disease after a 3 years and 6 months follow-up.
therapy and a radiation boost to the tumour bed. Local re-excision was not performed, since the initial excision already reached the infra-mammary fold. Patient 2. Local relapse was detected in the same lower-outer quadrant of a patient treated 33 months earlier by superior pedicle inverted-T plasty with local free margins superior to 10 mm for a large DCIS. She underwent whole breast radiation therapy without radiation boost to the tumour bed.
Patient 3. Local recurrence occurred after lateral mammaplasty for a 30 mm DCIS, with clear histologic margins. The patient underwent whole breast radiation therapy and a radiation boost to the tumour bed. Twenty-five months later, invasive local recurrence in the same upper-outer quadrant was revealed by the appearance of new calcifications at mammography. Patient 4. A ductal invasive carcinoma was detected in the same upper-outer quadrant of a high-grade DCIS previously treated 21
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
4
C. Malhaire et al. / The Breast xxx (2015) 1e6
Table 1 Pre-operative mammographic findings.
Calcifications type Round Amorphous Coarse heterogeneous Fine linear Calcifications distribution Grouped Linear Segmental Regional Breast density BIRADS A BIRADS B BIRADS C BIRADS D
Total
Positive margins
Negative margins (>1 mm)
4 (3.6%) 11 (9.8%) 77 (68.8%) 20 (17.9%)
2 (5%) 4 (9%) 30 (70%) 7 (16%)
2 (3%) 7 (10%) 47 (68%) 13 (19%)
14 20 62 14
(12.7%) (18.2%) (56.4%) (12.7%)
4 (9%) 8 (19%) 28 (65%) 3 (7%)
10 12 34 11
20 (18.9%) 46 (43.4%) 35 (33.0%) 5 (4.7%)
10 (24%) 16 (38%) 14 (33%) 2 (5%)
10 (16%) 30 (47%) 21 (33%) 3 (5%)
p value p 0.53
p 0.33 (15%) (18%) (51%) (16%) p 0.72
Table 2 Pre-operative histologic diagnosis. Histologic results of pre-operative biopsy
N
DCIS Micro-invasive Invasive DCIS Tumor grade Grade: low Grade: intermediate Grade: high Non evaluable Invasive carcinoma Eston Ellis grade Grade 1 Grade 2 Grade 3 Non evaluable Missing data Receptor status ERþ ER Not performed Missing data PRþ PR Not performed Missing data cErbB2 status Not performed Non overexpressed/þ þþ þþþ Missing data Mitotic index Not performed Low Intermediate High Missing data
80 9 24 5 37 36 2
months earlier by lateral mammaplasty with a 3 mm minimal margin and whole breast radiation therapy with a radiation boost to the tumour bed. This mammographically occult lesion was shown by breast ultrasound, prompted by the appearance of an axillary metastatic lymph node. Patient 5. A multifocal invasive ductal carcinoma occurred at the union of outer quadrants 79 months after OS of an upper outer microinvasive ductal carcinoma resected with clear histologic margins, followed by whole breast radiation therapy and a radiation boost to the tumour bed. Discussion
4 12 4 4 2 12 4 3 12 4 5 3 5 4 9 5 0 6 6 7 3 3 5
Our results indicate that OS following BWL allowed the resection of significantly larger lesions than SL, while maintaining equivalent rates of clear histologic margins. Hence, while reoperation rates compares to SL, OS following BWL enabled the resection of much larger lesions with comparable success of calcifications removal at post-operative mammogram. The median mammographic size of 47 mm in our study is, to our knowledge, the largest reported after BWL, while the volume of tissue resected reached a median value of 172 cc3 [range 180e1980 cc3, 298 cc3 mean], allowing for clear histologic margins in 61% of the patients. Results are difficult to compare with previous studies, given the difference in inclusion criteria and definition of positive margins, which range from the presence of tumoural cells on ink borders to 3 mm. Yet our results compare favorably with previous series dealing with BWL before BCS [8,14,15], which are summarized in Table 5. In a prospective non-randomized study, Liberman et al. reported results after BWL for 103 lesions including benign, high risk and malignant lesions. In the malignant group, clear histologic margins were achieved in 44%. Notably, the only predictor of positive margin status was breast density [15]. Absence of residual calcifications was achieved in 74% of the bracketed cancers, compared to 76% in our series. In this study, the likelihood of obtaining clear histologic margins was significantly higher if calcifications were completely removed than Table 4 Oncoplastic versus standard breast-conserving surgery.* Percentages calculated from 75 patients with post-operative mammogram available.
Table 3 Summary of oncoplastic techniques used in the study. Oncoplastic technique
N ¼ 73
Lateral mammaplasty Inverted-T (superior pedicle) Omega J-plasty Inverted-T (inferior pedicle) Peri-areolar Infra-mammary fold Medial mammaplasty
37 15 5 4 4 3 1 4
Standard lumpectomy
Oncoplasty
N
40 (35%)
73 (65%)
Volume (cc3, median) Negative margins No residual calcifications* Re-operation Local re-excision Mastectomy
88 25/40 (63%) 18/25 (72%) 16/40 (40%) 9/16 (56%) 7/16 (43%)
246 44/73 (60%) 39/50 (78%) 31/73 (42%) 6/31 (20%) 25/31 (80%)
p value
p p p p p
< 0.0001 ¼ 0.97 ¼ 0.33 ¼ 0.42 ¼ 0.01
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
NS NS 23% 10% 33% 12% 24% 25% 28% 36% 44% 25% 44 g 31 g 85 ml 53 ml 188 ml 137 ml NS NS 77% 90% 68% 82% 26 12.5 NS NS NS NS NS Overall 28 (mean) Malignant 79 Fillion 2012 [14]
100%
NS Malignant 100% 153 Kirstein 2008 [8]
37 Cordiner 2006 [16]
NS
Yes 37 BWL 101 SWL
Contents lists available at ScienceDirect
The Breast journal homepage: www.elsevier.com/brst
Original article
Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer Caroline Malhaire a, *, Delphine Hequet b, Marie-Christine Falcou c, nie Guillot b, Jean-Guillaume Feron b, Anne Tardivon a, Alexandre Leduey b, Euge c b b ronique Mosseri , Roman Rouzier , Benoit Couturaud , Fabien Reyal b Ve a b c
Department of Radiology, Institut Curie, 26 rue d’Ulm, 75005 Paris, France Department of Surgical Oncology, Institut Curie, 26 rue d’Ulm, 75005 Paris, France Department of Public Health, Institut Curie, 26 rue d’Ulm, 75005 Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 November 2014 Received in revised form 16 February 2015 Accepted 22 February 2015 Available online xxx
Purpose: The purpose of this study was to evaluate the outcome of breast conserving surgery comparing oncoplastic surgery (OS) and standard lumpectomy (SL) after preoperative bracketing wire localization of large neoplastic lesions. Methods: We retrospectively reviewed the medical records and the mammograms of patients operated on at the Institut Curie between May 2005 and September 2011 after bracketing wire localization under mammographic and/or sonographic guidance. Results: 113 patients underwent surgery for a pre-operative diagnosis of DCIS (n ¼ 80), micro-invasive carcinoma (n ¼ 9) or invasive carcinoma (n ¼ 24), by OS (n ¼ 73) or SL (n ¼ 40). In the OS group, radiological size (52 mm vs 39 mm, p < 0.001) and resection volumes (246 cc vs 88 cc, p < 0.00001) were significantly higher than in the SL group. Rates of clear histologic margins (60 vs 62%, NS), complete excision of microcalcifications (78% vs 72%, NS) and re-intervention rate (40% vs 42%, NS) were equivalent. The rate of local recurrence at 24 months was 3% [0e7.1] in patients with conservative treatment (n ¼ 3). With a median follow-up of 40 months, 5 local relapses (two with axillary metastatic involvement), two distant metastatic evolution, one contralateral breast cancer and one death unrelated to cancer occurred. Conclusion: Following bracketing wire localization, OS allowed the conserving management of significantly larger lesions with wider resection volumes, without significant increase in margin involvement or re-intervention rate, and equivalent rate of microcalcifications clearance compared to SL. © 2015 Elsevier Ltd. All rights reserved.
Introduction Breast-conservation surgery (BCS) allows equivalent survival outcome to that of mastectomy, although the rate of local recurrence is slightly higher [1,2]. BCS involves performing a resection wide enough to ensure margins clearance and local control, while maintaining acceptable cosmetic outcome. Despite this, poor cosmetic outcomes are still reported in up to 40% of patients after BCS [3]. Rather than an absolute tumour dimension, ratio of tumour-to-breast and tumour location are determinant keys for the indication of conservative surgery.
* Corresponding author. Tel.: þ33 1 44 32 42 81. E-mail address: [email protected] (C. Malhaire).
Recent years have seen the emergence of oncoplastic techniques combining plastic surgery procedures of the breast to standard lumpectomy. Oncoplastic surgery (OS) relies on advancement, rotation or transposition of a breast-flap [4], allowing for a resection of lesions that would otherwise not be amenable to standard lumpectomy, while also allowing for better cosmetic results [5]. Large calcifications clusters revealing breast cancer is a challenging situation regarding breast conservation. Extensive ductal carcinoma in situ (DCIS) is frequently growing in a radial fashion [6], as a contiguous disease in the same segment of the breast [7] and may thereby benefit from oncoplastic techniques, as it is not appropriate for standard lumpectomy (SL). Multiple wire localization techniques have been recommended to bracket large calcifications clusters before performing BCS in order to better document
http://dx.doi.org/10.1016/j.breast.2015.02.037 0960-9776/© 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
2
C. Malhaire et al. / The Breast xxx (2015) 1e6
lesion extension [8]. The aim of our study was to assess the outcome of BCS with wide local excision of breast carcinoma using bracketing wire localization (BWL), regarding complete excision of calcifications at post-operative mammogram, margin status, reoperation and completion mastectomy rates. Materials and methods Local ethical committee approved retrospective review of patients' records and imaging studies. From our radiological report data of May 2005 to September 2011, we retrospectively identified 188 consecutive BWL performed under mammographic and/or sonographic guidance for breast carcinoma. Exclusion criteria included benign or atypical lesions (n ¼ 42), neoadjuvant chemotherapy (n ¼ 16), wire localization performed for distinct lesions (n ¼ 17). A total of 113 cases with a preoperative needle biopsy diagnosis of DCIS and/or invasive carcinoma were finally included. All mammograms were retrospectively reviewed by a radiologist trained to breast imaging (CM, 10 years experience) without knowledge of post-operative results. Breast density according to the BI-RADS classification, type and distribution of calcifications and two-dimensional size of the calcifications cluster were recorded using BI-RADS lexicon [9]. The maximal axis size of the lesion measured on each view was retained for analysis. BWL was performed under mammographic or sonographic guidance for lesions considered to be too large to be accurately marked by single wire localization. Usually, a lesion of 3 cm or more was regarded as an indication for bracketing wire localization in our institution [10]. Five or 9 cm length wires (Medi-tech/Boston Scientific, Watertown, Mass and Cook, Bloomington, IA) were placed in an attempt to mark out the lesion and define as precisely as possible the upperlower and latero-medial extension. As much as possible, wire needles were inserted from the same side of the breast with the shortest approach to facilitate surgical dissection. Wire localization was performed using a compression paddle with an opening area (Mammomat 3000, Siemens, Erlangen Germany and Senograph Essential, GE Healthcare, Milwaukee, Wisconsin). Latero-medial and cranio-caudal mammograms were systematically performed after wire deployment to assess the position of wires relative to the calcifications cluster (Fig. 1AeC). A metallic dot was placed to help surgeons identify the nipple. After resection, a radiograph of the oriented specimen was performed to document successful removal of breast calcifications and to assess radiographic margins (Fig. 1D). After resection of large calcifications clusters, post-operative mammogram was recommended [11]. The largest dimension of the lesion was determined from pathology reports and volume was calculated using the specimen's length, height and width. Clear histological margins were defined as superior to 1 mm [12]. Date and type of secondary surgery and histological results were recorded, as well as follow-up medical records of local or distant relapse, and second cancers. OS was performed by either plastic surgeons, or breast surgeons with OS training. Statistical analysis Categorical variables were summarized as counts and percentages in each class, the quantitative variables as mean and standard deviation, median, minimum, maximum. The observed percentages were compared by the chi-square test or Fisher's exact test. Medians were compared using the nonparametric Kruskal Wallis test. The differences were considered significant if p is less than 0.05.
Results The median age was 58 years [range 38e63]. Four patients had high familial risk according to genetic counseling, without mutation identification in three of them. Tests results were still pending for the fourth patient. Median maximal dimension of calcifications clusters measured on preoperative mammograms was 47 mm overall [range, 15e115 mm], significantly larger in oncoplastic group (median 52 mm, range: 15e92) than in SL group (median 39 mm, range: 19e115, p < 0.01). In 7 patients, pre-operative mammograms were not available to assess calcifications cluster size. None of the radiographic features reported correlated with positive margins (Table 1). Pre-operative biopsy (Table 2) showed DCIS in 80, microinvasive DCIS in 9 and invasive carcinoma in 24. Of 113 patients, 73 underwent OS. Surgical techniques used during OS are summarized in Table 3 [13]. Comparative results of SL and OS are summarized in Table 4. The median resection volume (not available in 17 cases) was 172 cc3 overall [range 180e1980 cc3, mean 298 cc3, standard deviation 385]. significantly larger in the oncoplastic group with a median of 246 cc3 [range 42e1980 cc3] versus a median of 88.5 cc3 [range 18e448 cc3] in the SL group (p < 0.0001). Seventeen out of 80 lesions (21%) diagnosed as DCIS at percutaneous biopsy appeared to be invasive at final operative histologic analysis, while 6 of the 9 micro-invasive carcinomas were found to be invasive at surgical histologic analysis. Median size was 9 mm [range 2e120 mm] for invasive carcinomas and 35 mm [range 6e100 mm] for DCIS. Clear histologic margins were obtained in 61% with a median margin of 3.5 mm [range 1e20 mm, mean 4 mm]. In the case of positive margins, extent of margin involvement (available in n ¼ 40) was focal (15 mm) in 3 and multifocal in 12. Lesions resected with positive margins had a median mammographic size of 50 mm [range 19e115 mm], whereas those with negative margins had a median mammographic size of 46 mm [range 15e92 mm] (NS, p ¼ 0.15, Student t test). Margins status was not significantly associated with type of surgery (p ¼ 0.82, chi-square), as shown in Table 4. Fifty-seven of 75 patients with post-operative mammograms (76%) did not show any residual calcifications. Twenty patients underwent immediate local re-excision because of positive margins at histological analysis without post-operative mammogram. Forty-seven patients underwent local re-excision by BCS (n ¼ 15) or mastectomy (n ¼ 32) about half of which associated with immediate reconstruction. Overall, 31 patients had no post-operative radiotherapy. 26 had post-operative radiotherapy without local boost radiation and 56 patients underwent boost radiation following surgery, of whom 4 had close or positive margins, and 52 had final clear histologic margins (considering eventual local re-excision). The rate of local recurrence at 24 months was 3% [0e7.1] in patients with BCS. During a median of 40 months' follow-up assessment [1e91 months], two patients died from metastatic evolution, and one death was unrelated to cancer evolution. Local relapses, detailed below, occurred in 5 patients in the same quadrant as index cancer among the 81 patients who underwent BCS (6%), of whom two had axillary metastatic involvement. No relapse was recorded in the mastectomy group. One contralateral breast cancer was recorded. Patient 1. Local relapse occurred in the same lower-outer quadrant 26 months after the calcifications cluster had been bracketed and the patient operated on by quadrantectomy. Inferior margin was focally (5 mm) less than 1 mm for invasive and noninvasive carcinoma. She underwent whole breast radiation
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
C. Malhaire et al. / The Breast xxx (2015) 1e6
3
Fig. 1. 52-year-old woman with a preoperative diagnosis of IDC and DCIS from the US-guided microbiopsy of a hypoechoic nodule, associated with extensive calcifications at mammography. Microcalcification cluster extended 5 cm antero-posteriorly and 6 cm large on the cranio-caudal magnification view (A). Cranio-caudal view (B) and medio-lateral view (C) show BWL, one posterior wire indicating lateral and upper limit of the microcalcification cluster and another anterior wire indicating inner and lower limit of the microcalcification cluster. A J-plasty OS was performed allowing the resection of a 146 g specimen containing calcifications in its center at intraoperative radiograph (D). Wires are visible in the specimen, delineating lesion borders. No residual calcification of the large cluster was visible on cranio-caudal at postoperative mammogram. Histological analysis showed bifocal invasive ductal carcinoma of 18 and 5 mm (grade II, low mitotic index) associated with extensive DCIS. Minimal interval to lateral margins was 3 mm. Patient underwent chemotherapy, radiotherapy of the breast with local boost complement, radiation of the internal mammary, sus and sub clavian nodes and hormonotherapy (Aromasine®). She is free from disease after a 3 years and 6 months follow-up.
therapy and a radiation boost to the tumour bed. Local re-excision was not performed, since the initial excision already reached the infra-mammary fold. Patient 2. Local relapse was detected in the same lower-outer quadrant of a patient treated 33 months earlier by superior pedicle inverted-T plasty with local free margins superior to 10 mm for a large DCIS. She underwent whole breast radiation therapy without radiation boost to the tumour bed.
Patient 3. Local recurrence occurred after lateral mammaplasty for a 30 mm DCIS, with clear histologic margins. The patient underwent whole breast radiation therapy and a radiation boost to the tumour bed. Twenty-five months later, invasive local recurrence in the same upper-outer quadrant was revealed by the appearance of new calcifications at mammography. Patient 4. A ductal invasive carcinoma was detected in the same upper-outer quadrant of a high-grade DCIS previously treated 21
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
4
C. Malhaire et al. / The Breast xxx (2015) 1e6
Table 1 Pre-operative mammographic findings.
Calcifications type Round Amorphous Coarse heterogeneous Fine linear Calcifications distribution Grouped Linear Segmental Regional Breast density BIRADS A BIRADS B BIRADS C BIRADS D
Total
Positive margins
Negative margins (>1 mm)
4 (3.6%) 11 (9.8%) 77 (68.8%) 20 (17.9%)
2 (5%) 4 (9%) 30 (70%) 7 (16%)
2 (3%) 7 (10%) 47 (68%) 13 (19%)
14 20 62 14
(12.7%) (18.2%) (56.4%) (12.7%)
4 (9%) 8 (19%) 28 (65%) 3 (7%)
10 12 34 11
20 (18.9%) 46 (43.4%) 35 (33.0%) 5 (4.7%)
10 (24%) 16 (38%) 14 (33%) 2 (5%)
10 (16%) 30 (47%) 21 (33%) 3 (5%)
p value p 0.53
p 0.33 (15%) (18%) (51%) (16%) p 0.72
Table 2 Pre-operative histologic diagnosis. Histologic results of pre-operative biopsy
N
DCIS Micro-invasive Invasive DCIS Tumor grade Grade: low Grade: intermediate Grade: high Non evaluable Invasive carcinoma Eston Ellis grade Grade 1 Grade 2 Grade 3 Non evaluable Missing data Receptor status ERþ ER Not performed Missing data PRþ PR Not performed Missing data cErbB2 status Not performed Non overexpressed/þ þþ þþþ Missing data Mitotic index Not performed Low Intermediate High Missing data
80 9 24 5 37 36 2
months earlier by lateral mammaplasty with a 3 mm minimal margin and whole breast radiation therapy with a radiation boost to the tumour bed. This mammographically occult lesion was shown by breast ultrasound, prompted by the appearance of an axillary metastatic lymph node. Patient 5. A multifocal invasive ductal carcinoma occurred at the union of outer quadrants 79 months after OS of an upper outer microinvasive ductal carcinoma resected with clear histologic margins, followed by whole breast radiation therapy and a radiation boost to the tumour bed. Discussion
4 12 4 4 2 12 4 3 12 4 5 3 5 4 9 5 0 6 6 7 3 3 5
Our results indicate that OS following BWL allowed the resection of significantly larger lesions than SL, while maintaining equivalent rates of clear histologic margins. Hence, while reoperation rates compares to SL, OS following BWL enabled the resection of much larger lesions with comparable success of calcifications removal at post-operative mammogram. The median mammographic size of 47 mm in our study is, to our knowledge, the largest reported after BWL, while the volume of tissue resected reached a median value of 172 cc3 [range 180e1980 cc3, 298 cc3 mean], allowing for clear histologic margins in 61% of the patients. Results are difficult to compare with previous studies, given the difference in inclusion criteria and definition of positive margins, which range from the presence of tumoural cells on ink borders to 3 mm. Yet our results compare favorably with previous series dealing with BWL before BCS [8,14,15], which are summarized in Table 5. In a prospective non-randomized study, Liberman et al. reported results after BWL for 103 lesions including benign, high risk and malignant lesions. In the malignant group, clear histologic margins were achieved in 44%. Notably, the only predictor of positive margin status was breast density [15]. Absence of residual calcifications was achieved in 74% of the bracketed cancers, compared to 76% in our series. In this study, the likelihood of obtaining clear histologic margins was significantly higher if calcifications were completely removed than Table 4 Oncoplastic versus standard breast-conserving surgery.* Percentages calculated from 75 patients with post-operative mammogram available.
Table 3 Summary of oncoplastic techniques used in the study. Oncoplastic technique
N ¼ 73
Lateral mammaplasty Inverted-T (superior pedicle) Omega J-plasty Inverted-T (inferior pedicle) Peri-areolar Infra-mammary fold Medial mammaplasty
37 15 5 4 4 3 1 4
Standard lumpectomy
Oncoplasty
N
40 (35%)
73 (65%)
Volume (cc3, median) Negative margins No residual calcifications* Re-operation Local re-excision Mastectomy
88 25/40 (63%) 18/25 (72%) 16/40 (40%) 9/16 (56%) 7/16 (43%)
246 44/73 (60%) 39/50 (78%) 31/73 (42%) 6/31 (20%) 25/31 (80%)
p value
p p p p p
< 0.0001 ¼ 0.97 ¼ 0.33 ¼ 0.42 ¼ 0.01
Please cite this article in press as: Malhaire C, et al., Outcome of oncoplastic breast-conserving surgery following bracketing wire localization for large breast cancer, The Breast (2015), http://dx.doi.org/10.1016/j.breast.2015.02.037
NS NS 23% 10% 33% 12% 24% 25% 28% 36% 44% 25% 44 g 31 g 85 ml 53 ml 188 ml 137 ml NS NS 77% 90% 68% 82% 26 12.5 NS NS NS NS NS Overall 28 (mean) Malignant 79 Fillion 2012 [14]
100%
NS Malignant 100% 153 Kirstein 2008 [8]
37 Cordiner 2006 [16]
NS
Yes 37 BWL 101 SWL
