Inclusion of patients from the Danish National Patient Registry from January 1, 2010, through December 31, 2013. BCS indicates breast-conserving surgery; DBCG, the Danish Breast Cancer Group; DCIS, ductal carcinoma in situ; DNPR, the Danish National Patient Registry; DPR, the Danish Pathology Registry; and IBC, invasive breast cancer.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Langhans L, Jensen M, Talman MM, Vejborg I, Kroman N, Tvedskov TF. Reoperation Rates in Ductal Carcinoma In Situ vs Invasive Breast Cancer After Wire-Guided Breast-Conserving Surgery. JAMA Surg. 2017;152(4):378–384. doi:10.1001/jamasurg.2016.4751
What is the reoperation rate in patients with verified nonpalpable invasive breast cancer or ductal carcinoma in situ treated with wire-guided breast-conserving surgery?
In this population-based registry study, of 4118 patients who underwent wire-guided breast-conserving surgery, the overall reoperation rate was 17.6% (14.4% reexcisions and 3.2% mastectomies). The risk of a reoperation owing to positive margins was 3 times higher in patients with ductal carcinoma in situ (37.3%) vs those with invasive breast cancer (13.4%).
The reoperation rate for invasive breast cancer was lower than anticipated; however, the risk of reoperation in patients with ductal carcinoma in situ is still high and a more accurate localization method is needed.
New techniques for preoperative localization of nonpalpable breast lesions may decrease the reoperation rate in breast-conserving surgery (BCS) compared with rates after surgery with the standard wire-guided localization. However, a valid reoperation rate for this procedure needs to be established for comparison, as previous studies on this procedure include a variety of malignant and benign breast lesions.
To determine the reoperation rate after wire-guided BCS in patients with histologically verified nonpalpable invasive breast cancer (IBC) or ductal carcinoma in situ (DCIS) and to examine whether the risk of reoperation is associated with DCIS or histologic type of the IBC.
Design, Setting, and Participants
This nationwide study including women with histologically verified IBC or DCIS having wire-guided BCS performed between January 1, 2010, and December 31, 2013, used data from the Danish National Patient Registry that were cross-checked with the Danish Breast Cancer Group database and the Danish Pathology Register.
Main Outcomes and Measures
Reoperation rate after wire-guided BCS in patients with IBC or DCIS.
Wire-guided BCS was performed in 4118 women (mean [SD] age, 60.9 [8.7] years). A total of 725 patients (17.6%) underwent a reoperation: 593 were reexcisions (14.4%) and 132 were mastectomies (3.2%). Significantly more patients with DCIS (271 of 727 [37.3%]) than with IBC (454 of 3391 [13.4%]) underwent a reoperation (adjusted odds ratio, 3.82; 95% CI, 3.19-4.58; P < .001). After the first reexcision, positive margins were still present in 97 patients (16.4%). The risk of repeated positive margins was significantly higher in patients with DCIS vs those with IBC (unadjusted odds ratio, 2.21; 95% CI, 1.42-3.43; P < .001). The risk of reoperation was significantly increased in patients with lobular carcinoma vs those with ductal carcinoma (adjusted odds ratio, 1.44; 95% CI 1.06-1.95; P = .02). A total of 202 patients (4.9%) had a subsequent completion mastectomy, but no difference was found in the type of reoperation between patients with DCIS and those with IBC.
Conclusions and Relevance
A lower reoperation rate after wire-guided BCS was found in this study than those shown in previous studies. However, the risk of reoperation in patients with DCIS was 3 times higher than in those with IBC. The widespread use of mammographic screening will increase the number of patients diagnosed with DCIS, making a precise localization of nonpalpable DCIS lesions even more important.
The introduction of national mammographic screening programs has led to an increased number of identified nonpalpable breast lesions, and, accordingly, an increased need for preoperative localization of the lesions.1 The increase reflects detection of small invasive breast cancers (IBCs) and detection of ductal carcinoma in situ (DCIS), which was rarely detected before the introduction of mammographic screening owing to lack of clinical symptoms.2 Patients diagnosed with DCIS represent 15% to 20% of cancers detected via mammographic screening, with variations between countries.3,4 In Denmark, the national introduction of the organized mammographic screening program took place from 2007 to 2010.
Patients with nonpalpable lesions are often eligible for breast-conserving surgery (BCS) guided by preoperative localization of the lesions. It is important to remove all tumor tissue since positive margins are associated with local recurrence, and to maintain as much healthy breast tissue as possible to ensure an acceptable cosmetic outcome.5 The choice of BCS vs mastectomy depends on the extent of the lesion, size of the breast, patient preference, and comorbidities. Breast-conserving surgery combined with postoperative radiotherapy and mastectomy leads to similar rates of survival in patients with IBC, provided that the margins are free of tumor.6
The standard method for localization of nonpalpable breast lesions is wire-guided localization.7 However, several disadvantages of wire-guided localization are described in the literature; the most important one being the relatively large amount of patients with positive microscopic margins detected at the final pathologic evaluation.8-10 Reoperation with further resection is required when a positive margin is present; such a procedure can be both physically and mentally exhausting for the patient and can impair the cosmetic outcome.11 In addition, reoperation delays adjuvant treatment and is an extra financial expense. The percentages of patients with positive margins after wire-guided BCS reported in the literature vary largely, from 14% to 70%.8-10 New techniques for preoperative localization may decrease the risk of reoperation, but a valid rate of reoperation for the wire-guided procedure needs to be established.
The margins of DCIS lesions are less well defined than those of IBC lesions, and the extension into the breast tissue can be difficult to determine.12,13 Accordingly, a higher rate of reoperation in patients with DCIS than in those with IBC is expected. To our knowledge, no study has examined rates of reoperation in patients with verified DCIS treated with wire-guided BCS.
We examined patients with verified nonpalpable IBC or DCIS treated with wire-guided BCS. Our aim was to estimate the total rate of reoperation, including the rate of reexcisions owing to positive margins. In addition, we examined whether the risk of a reoperation or a subsequent completion mastectomy differed according to diagnosis, age, or histologic type of the invasive carcinoma.
Data on all women with IBC or DCIS who underwent wire-guided BCS from January 1, 2010, through December 31, 2013, were retrieved from the Danish National Patient Registry. Patients are registered according to the diagnosis obtained at the final pathologic evaluation. Breast-conserving surgery is not indicated in patients with multifocal disease. Information on reexcision owing to positive margins within 2 months after the primary operation was also retrieved from the registry. The Danish National Patient Registry is a comprehensive register established in 1975 that contains data on all contacts with the Danish health care system, including admission to or examination at the hospital, diagnosis, treatment, and surgery. The Danish National Patient Registry is considered complete and has been quality assured.14
In January 2010, the registration of breast cancer surgery was standardized by the Danish Health and Medicines Authority and the Danish Society of Breast Surgery. The subsequent registration can therefore be considered consistent.
Information on patients with IBC retrieved from the Danish National Patient Registry was cross-checked with the Danish Breast Cancer Group (DBCG) database,15 a clinical database established in 1977. All departments involved in the treatment of breast cancer report information regarding diagnosis, treatment, and follow-up to the DBCG database. All Danish patients with IBC or DCIS are treated at departments certified for breast cancer treatment. The preoperative diagnosis is a variable contained in the DBCG database, which made it possible to ensure that patients who underwent wire-guided diagnostic excisional biopsies were excluded from the cohort. In addition, information on mastectomies performed within 2 months of the primary wire-guided BCS was retrieved from the DBCG database for patients with IBC and those with DCIS (Figure). Patients in our study were identified mainly through screening and referred for diagnostic mammography, including clinical examination and whole-breast ultrasonography. Preoperative magnetic resonance imaging is not standard practice in Denmark.
All patients registered with DCIS were cross-checked with the Danish Pathology Registry to ensure that a preoperative diagnosis of DCIS was present, since these data were not available through the DBCG database.16
In Denmark, margin status is defined according to the guidelines from the DBCG.17 A microscopic examination of representative sections from the resection margins is performed, and both IBC and DCIS require a microscopic negative margin of 2 mm at the final pathologic evaluation. If these criteria are not met, the margin is defined as positive. Lobular carcinoma in situ at the margin, except the pleomorphic type, is not an indication for additional surgery. In our study, reoperation contains both reexcisions and mastectomies. Reexcision was defined as that owing to positive margins, regardless of the number of margins resected. All patients registered with a reexcision were checked in the Danish Pathology Register and information regarding margin status after the reexcision and further surgery was obtained.
The study was approved by the National Committee on Health Research Ethics and the Danish Data Protection Agency, who did not require patient consent. All patient data were deidentified.
The rate of reoperations owing to positive margins was estimated as the percentage of patients registered with a reoperation out of the total number of patients undergoing wire-guided BCS during the 4-year study period. Differences in the rate of reoperation and in margin status after the first reexcision according to age and diagnosis were determined using a χ2 test. The need for further treatment between patients with DCIS and those with IBC was also tested using a χ2 test. Univariate and multivariable logistic regression were used to estimate odds ratios (ORs) with 95% CIs for reoperation and repeated positive margins, respectively. P < .05 was considered statistically significant. All statistical analyses were performed using SAS statistical software, version 5.1 (SAS Institute Inc).
A total of 4500 women with IBC or DCIS undergoing wire-guided BCS during the 4-year study period were identified in the Danish National Patient Registry. After cross-checking with the DBCG database and the Danish Pathology Register, 382 patients were excluded owing to procedure registration errors, mainly diagnostic excisional biopsies registered as lumpectomies, leaving 4118 patients for further analysis (Figure).
Of the 4118 patients included, 3391 were diagnosed with IBC and 727 were diagnosed with DCIS (Table 1). A total of 725 women (17.6%) underwent reoperation within 2 months of the primary wire-guided BCS; 593 (14.4%) had reexcision owing to positive margins and 132 (3.2%) had a mastectomy (Table 2).
Of the 727 patients diagnosed with DCIS, 271 (37.3%) underwent a reoperation compared with only 454 of the 3391 patients with IBC (13.4%) (Table 1) (adjusted OR, 3.82; 95% CI, 3.19-4.58; P < .001) (Table 3). Similar results were found by excluding the 132 patients having a mastectomy and examining only those who underwent reexcisions.
Of the 725 patients who had a reoperation performed, 454 had IBC (Table 1). These patients were divided into 3 groups: those with ductal carcinoma, lobular carcinoma, or other types of carcinoma. A significantly increased risk of reoperation was seen in patients with lobular carcinomas compared with patients with ductal carcinomas (adjusted OR, 1.44; 95% CI, 1.06-1.95; P = .02) (Table 3).
Of the 593 patients who had a reexcision performed, 97 (16.4%) were again found to have positive margins (Table 4). Of the 228 patients with DCIS, 53 (23.2%) had positive margins compared with 44 (12.1%) of the 365 patients with IBC. Patients with DCIS had a significantly higher risk of repeated positive margins compared with patients with IBC (unadjusted OR, 2.21; 95% CI, 1.42-3.43; P < .001) (Table 3). There was no statistically significant difference in the number of patients with invasive ductal carcinoma (35 of 292 [12.0%]) and with lobular carcinoma (7 of 46 [15.2%]) who had repeated positive margins (P = .54). A total of 89 patients underwent a second reoperation: 19 reexcisions (21.3%) and 70 mastectomies (78.7%). The remaining 8 patients did not receive additional surgical treatment despite having positive margins (Figure).
A total of 202 patients underwent mastectomy either as the first or second reoperation (Table 2). The remaining 523 patients underwent 1 or 2 reexcisions. There was no difference between patients with DCIS and patients with IBC in the type of reoperation. A total of 77 patients with DCIS (28.4%) underwent mastectomy, compared with 125 patients with IBC (27.5%) (unadjusted OR, 1.04; 95% CI, 0.75-1.46; P = .80). There was also no difference when comparing patients with ductal carcinomas and those with lobular carcinomas: 97 patients with ductal carcinoma (27.0%) and 18 patients with lobular carcinoma (31.0%) underwent mastectomy (unadjusted OR, 1.2; 95% CI, 0.67-2.20; P = .52).
The mean (SD) age of patients in the study was 60.9 (8.7) years. Patients were grouped in 4 categories to examine age as a risk factor for reoperation. Our analyses revealed that age younger than 50 years was associated with higher risk of reoperation after adjusting for DCIS vs IBC (OR, 1.86; 95% CI, 1.26-2.74) (Table 3). An interaction was seen between age and diagnosis: patients with IBC had a more clear decrease in rate of reoperations with increasing age compared with those with DCIS.
We found a total reoperation rate of 17.6% and a reexcision rate of 14.4% in patients with nonpalpable IBC or DCIS treated with wire-guided BCS. Patients with DCIS had a risk of reoperation 3 times that of patients with IBC. In addition, the risk of repeated positive margins was also significantly increased in patients with DCIS. The risk of undergoing a reoperation was significantly increased in patients with lobular carcinoma vs those with ductal carcinoma, which is in accordance with results of previous studies.18,19 There was no significant difference in the proportion of patients with DCIS who underwent a subsequent completion mastectomy vs patients with IBC. Young age has been shown to be associated with repeated surgery.12,18-20 This finding was also demonstrated in our study, even though patients in our study were diagnosed mainly through the national screening program and therefore are aged 50 to 69 years.
A recent study, including 316 114 patients from the National Cancer Data Base in the United States reported a higher reoperation rate than seen in our study.19 They concluded that approximately one-fourth of all patients who undergo initial BCS for IBC or DCIS will have a reoperation performed. A large cohort study from England demonstrated that 1 in 5 women who had BCS had a reoperation.12 These 2 studies do not take into account whether the lesion is nonpalpable nor do they outline whether preoperative localization was used. In some studies, the rate of reoperation after wire-guided BCS ranged from 14% to 70%.8-10 However, these figures are based on heterogeneous patient populations that have both benign and malignant lesions. In addition, the use of preoperative localization is not consistently reported.21-27 If diagnostic excisional biopsies are included, the rate of reoperation will presumably increase since the purpose of the primary surgery is to establish a diagnosis and not to perform radical surgery. This hypothesis is also supported by a Dutch study in which patients had a much higher risk of reexcision if a preoperative diagnosis was lacking.28 To our knowledge, our study is the largest to establish rates of reoperation in nonpalpable malignant breast lesions for which radical cancer surgery was intended. Patients in Denmark are treated according to the national guidelines from DBCG,17 which cover all aspects of diagnosis and treatment, including the assessment of pathologic margins. The treatment is centralized in a few departments with a high volume of patients with breast cancer. These factors could explain why the rate of reexcision is lower than reported in previous studies.
It is generally believed that it is more difficult to achieve negative margins in nonpalpable lesions owing to technical difficulties of wire-guided excisions and the diversity in radiographic appearance.9,13,22 To our knowledge, this view is argued by only 1 study, in which no significant difference in the rate of positive margins between palpable and nonpalpable lesions was seen.24 However, DCIS was excluded from that study, which might be the explanation for this result. The margins of DCIS lesions are less well defined and the extension into the breast is difficult to determine, which probably accounts for the high rate of reoperations seen in patients with DCIS vs those with IBC.12,28 In addition, an intraoperative macroscopic evaluation of resection margins is often not possible in DCIS lesions. A large retrospective study including both palpable and nonpalpable lesions found that a reoperation owing to positive margins was nearly twice as likely when the tumor had a carcinoma in situ component, compared with IBC.12 These results are confirmed in our study, where we found an even larger difference in the risk of reoperation, with an adjusted OR of 3.82 for reoperation in patients with DCIS vs those with IBC (Table 3).
Differences in guidelines for margin assessment and in the indication for a reoperation might influence the rates of reoperation. The definition of positive margin has been debated extensively the past decade and no universal definition exists. It is clear that a positive margin is associated with local recurrences, but the question is what margin width results in the lowest number of local recurrences.29
In 2010, an international expert panel outlined its recommendations regarding margin status.30 They recommended that the definition tumor not touching ink constituted an adequate negative margin in patients with IBC. A 2-mm negative margin was recommended for DCIS. This definition is supported by the 2013 St. Gallen Consensus Conference.31 A large meta-analysis from 2014 concluded that ensuring negative margins contributes to reducing the risk of local recurrence, but the adoption of wider margins is unlikely to have additional benefit for long-term local control in patients with IBC, as long as the inked margin is microscopically negative.29 Despite these recommendations, the minimum surgical resection margin is still a controversial issue. It was recently suggested that reexcisions can be omitted in patients with focally positive resection margins.32
Variation in rates of reoperation between studies can partly be owing to the lack of a consistent definition of a positive margin and to variation in the use of intraoperative pathologic assessment of margins. Besides gross evaluation, several other methods of intraoperative margin assessment are available to reduce the need of reoperation, such as frozen section analysis or touch preparation cytology. It is documented that frozen section analysis can identify positive margins and results in low rates of reoperation, but it increases the duration of the surgical procedure and diversion of pathologist resources.33 According to the DBCG guidelines, negative margin is defined as a microscopic tumor-free margin of 2 mm for both IBC and DCIS.17 It is standard to perform intraoperative gross evaluation of the tumor specimen in patients with IBC and tumor-generating DCIS. If the margin is less than 5 mm, additional shaving of the surgical cavity is performed for the affected margin. These guidelines are nationwide and apply for all patients in our study. Accordingly, our results are based on a uniform definition of positive margins. It can be argued that a definition of negative margins as tumor not touching ink would have decreased the rate of reoperation for patients with IBC in our study and increased the difference in rates of reoperation between patients with DCIS and those with IBC even further.
A reoperation owing to inadequate margin status can postpone adjuvant treatment and result in a poorer cosmetic outcome.12 A reoperation can be both physically and mentally exhausting for the patient, and efforts should be made to minimize the rate of reoperations. The technical difficulties of the wire-guided localization procedure are considered a reason for not achieving microscopically negative margins. It is challenging for the surgeon to estimate the exact distance and direction of the wire within the breast, because the surgeon does not have a 3-dimensional perspective on the lesion.34 The wire can be displaced, leading to an inaccurate localization, or it can be transected during surgery.10 New techniques, such as radioactive seed localization and radio-guided occult lesion localization for preoperative localization of nonpalpable lesions, may decrease the risk of a reoperation by being more accurate than wire-guided localization.7,9,35 The results of our study can be used as a valid reference to compare rates of reoperation after new methods of localization with rates of reoperation after the wire-guided procedure.
The high rate of reoperation in patients with DCIS found in our study indicates that these patients in particular might benefit from new and alternative methods of localization. This group of patients is increasing with the more widespread use of mammographic screening, making a precise localization of nonpalpable DCIS lesions even more important.
Completion mastectomies were performed because necessary radicality could not be achieved by BCS, often owing to multiple lesions identified after surgery. A distinct subgroup of patients that could be allocated to undergo immediate mastectomy could not be identified in our study. Better preoperative localization methods would presumably not decrease the risk of mastectomies but better preoperative imaging could benefit this small percentage of patients.
The strength of this study is the nationwide design based on a close to complete national register. All patients in the cohort have been cross-checked with either the DBCG database or the Danish Pathology Register to further minimize registration errors. All departments involved in the treatment of IBC and DCIS follow the national guidelines from DBCG, which produce consistent data regarding preoperative imaging, diagnosis, preoperative lesion localization, surgical treatment, and pathologic findings. The lesions in our study are nonpalpable and most are most likely 1 cm or less. In DCIS, tumor size is only measurable in the few that are tumor generating; therefore tumor size is not taken into account in our study, even though large tumor size has been shown to be associated with repeated surgery.19 In addition, estrogen receptor status, node status, and tumor grade are not included in our analyses. Previous studies have shown that especially high tumor grade is associated with a higher risk of reoperation.19,36 This is a limitation in our study and more studies are needed to clarify the importance of these covariates in the risk of reexcision.
Patients with nonpalpable lesions are often eligible for BCS guided by preoperative lesion localization. It is important to obtain negative resection margins but without removing additional healthy breast tissue to ensure an acceptable cosmetic outcome. The rate of reexcision after wire-guided BCS in this nationwide study was lower than anticipated. However, the risk of a reoperation and a reexcision after wire-guided BCS was 3 times higher in patients with DCIS than in those with IBC. The number of patients with nonpalpable DCIS lesions is increasing with the more widespread use of mammographic screening, making a precise localization of nonpalpable DCIS lesions even more important.
Accepted for Publication: September 30, 2016.
Corresponding Author: Linnea Langhans, MD, Department of Plastic Surgery, Breast Surgery and Burns, Rigshospitalet, University of Copenhagen, Blegdamsvej 9, DK-2100 Copenhagen, Denmark (firstname.lastname@example.org).
Published Online: December 21, 2016. doi:10.1001/jamasurg.2016.4751
Author Contributions: Dr Langhans had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Langhans, Kroman, Tvedskov.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Langhans, Jensen.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Langhans, Jensen.
Obtained funding: Langhans, Kroman.
Administrative, technical, or material support: Langhans, Jensen, Talman, Kroman.
Study supervision: Langhans, Jensen, Vejborg, Kroman, Tvedskov.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was funded by the Danish Cancer Society and the Danish Cancer Research Foundation.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: