Breast-conserving therapy for breast carcinoma : Margins , re-excision and recurrence rates

Until the 1970s, the Halsted radical mastectomy had been the surgical procedure in the management of breast cancer irrespective of size. It was described and performed by William Stewart Halsted in 1882.3 It included the excision of the skin, the breast and underlying pectoralis muscles. Patients also underwent an axillary lymph node dissection. A skin graft was often used to close the wound defect. Patients were left with morbid disfiguring appearances.


Introduction
Breast cancer affects 1 million women in the developed world every year. 1 The burden of breast cancer is also on the rise in the developing world.According to Jones, 2 it is estimated that 70% of new cancer cases will occur in the developing world by the year 2020.
Until the 1970s, the Halsted radical mastectomy had been the surgical procedure in the management of breast cancer irrespective of size.It was described and performed by William Stewart Halsted in 1882. 3It included the excision of the skin, the breast and underlying pectoralis muscles.Patients also underwent an axillary lymph node dissection.A skin graft was often used to close the wound defect.Patients were left with morbid disfiguring appearances.
Modified radical mastectomy was shown to be as effectual as radical mastectomy and less mutilating.Modified radical mastectomy remains to be applicable to certain patients.However, breast-conserving therapy (BCT) has become the adopted procedure for the surgical management of early breast cancer.BCT refers to the surgical excision of the primary breast cancer tumour with a rim of normal breast tissue, followed by radiation therapy.It is also referred to in the literature as lumpectomy, partial mastectomy or segmental mastectomy. 4er the past 40 years, prospective large multicentre randomised controlled trials (RCTs) in Europe and United States have provided level 1 evidence that there is no statistically significant difference in the overall long-term survival between mastectomy and BCT in appropriately selected patients treated for early breast cancer. 5Although the overall long-term survival is similar between the two procedures, patients undergoing BCT are at a recognised risk of ipsilateral breast tumour recurrence. 6The success rate of BCT relies upon appropriate patient selection, surgical technique and postoperative radiotherapy to eliminate microscopic tumour cells in the ipsilateral breast.The goals of BCT are therefore to resect the tumour with clear microscopic margins and to achieve acceptable cosmetic outcomes without compromising patient survival.
Background: Breast-conserving therapy (BCT) is a wide local excision of the tumour usually followed by radiation treatment to the breast.It is the mainstay treatment for carefully selected patients with early breast cancer.There has not been a formal audit to review BCT outcomes in our unit.
Objectives: To determine excision margins, re-excision and local recurrence rates.
Methods: A histopathological and oncology records' review of BCT patients from 01 January 2006 to 31 December 2010.The health faculty's ethics committee granted approval.Data points accrued included age, histological tumour size, nodal status, tumour type, oestrogen receptor status, lymphovascular invasion, volume of specimen, margin status, management of involved or close margins, radiotherapy, ipsilateral breast recurrence rate and duration of follow-up.

Conclusion:
Positive and close margin re-excision and local recurrence rates in our unit are acceptable and comparable to other units in South Africa and internationally.
Breast-conserving therapy has become the standard of care in our oncology and endocrine surgical unit for patients meeting the above criteria diagnosed with early breast cancer at Groote Schuur Hospital.There has not been a formal audit to review the outcomes of BCT in the unit.In South Africa, there has been only one retrospective study on BCT.In 2005, Mannell, a part-time staff member at the University of Witwatersrand, published a retrospective review of 165 patients that underwent BCT at her private practice at Linksfield and Parklane Clinics over a period of 12 years.In her series, 7 out of 165 patients had re-excision and recurrence at a rate of 5.5%. 7

Materials and methods
This is a 5-year histopathological and oncology records' review of consecutive patients who underwent BCT in a specialised unit at Groote Schuur Hospital from 01 January 2006 to 31 December 2010.Inclusion criteria are all the histologically confirmed invasive or in situ breast carcinoma patients.Exclusion criteria are the patients with benign pathological diagnoses, patient with missing folder numbers, patients whose folders could not be found at records and patients lost to follow-up during the study period.Primary endpoints were margin status, rate and type of re-operation and recurrence rate.Secondary outcomes are to identify factors associated with margin status, re-excision and recurrence.Data were retrieved from the unit's prospectively collected patient surgical breast cancer Microsoft Access 2010 database by identifying all the patients that had wide local excisions (WLEs).Additional data were retrieved from patients' folders, National Health Laboratory Service (NHLS) pathology reports and the radiation oncology records.
Data from identified patients were captured using an online password-protected form designed with Google Forms and subsequently transferred onto a spreadsheet.Limiting access and using proxy patient identifiers maintained patient confidentiality.
The following variables were captured: age, pathological tumour size (pT), pathological node status (pN), tumour grade, histological type of tumour, volume of resected specimen (cm 3 ), oestrogen receptor (ER) status, presence of lymphovascular invasion, margin status, type of tumour at margin, management of positive margins (re-excision vs mastectomy), type of residual tumour in the resection or mastectomy specimen, radiation therapy post WLE -if completed or not -local recurrence (LR) and time to recurrence, type of tumour and management of recurrence, duration of follow-up at oncology outpatients' department.Extensive intraductal component (EIC) was not recorded, as it did not feature in the NHLS pathology reports.
Tumour grade is described as low, intermediate or high, which is equivalent to 1, 2 or 3, respectively, as per the Nottingham grading system (also called the Elston-Ellis modification of the Scarff-Bloom-Richardson grading system) for breast cancer.
Positive margin is interpreted as the presence of breast cancer, invasive or non-invasive at the inked surgical margin.The absence of tumour within a specified distance of more than 1 mm or 2 mm from the resection margin was regarded as a negative margin.A close margin is the presence of tumour within that distance (1 mm) but not at the resection margin.Volume of specimen is the gross volume of excised cancer with surrounding breast tissue calculated by multiplying height by width by length using the dimensions as described in the pathology report.
The wide local excision procedure in the unit is performed as per standards for BCT. 4 An elliptical skin incision is followed by excision of the tumour together with circumferential normal breast tissue of 1 cm to ensure that the resected specimen margins are free of the tumour with tactile perception.The axilla was managed as indicated by clinical or sentinel lymph node biopsy.Lymph node clearance was limited to Level II dissection.LR refers to the first site with histologically proven relapse of invasive or intraductal carcinoma post WLE in the ipsilateral breast.
Statistical analysis was performed using the IBM SPSS Statistics version 22 for Macintosh software.Continuous variables were expressed as means and median as measures of central tendencies and categorical variables as frequencies and percentages.The chi-square test was used to determine the associations between categorical data variables to recurrence.

Results
Table 1 shows patients and disease characteristics.
The query design for the unit's Microsoft Access database retrieved 242 patients that had procedures recorded as WLEs.Fifty patients were excluded from the study for the following reasons: lost to follow-up (31), incomplete histopathology report (8), benign pathology (3) (benign phylloides 2 and juvenile fibroadenoma 1), diagnostic radio-guided occult lesion localisation (2), incomplete database entry (3) (no name and folder number) and unable to retrieve pathology or oncology records (3).Therefore, a total of 192 patients underwent BCT in the surgical oncology unit between 01 January 2006 and 31 December 2010 as per inclusion criteria.The patients' age ranged from 25 years to 84 years with a mean age of 53.4 years.More than half of the patients (54.7%) were above 51 years; 5.2% of patients were young women who were < 35 years.The majority of patients (91.7%) had breast cancers with a maximal diameter of 5 cm.A total of five and four patients had tumours categorised as T3 and T4 by pathology, respectively.Seven patients had WLE post neoadjuvant chemotherapy.
The commonest histological tumour type was infiltrative ductal carcinoma (IDC) at 79.1%.It was associated with ductal carcinoma in situ (DCIS) in 40.6%.Infiltrative lobular carcinoma (ILC) was present in 5.3% followed by DCIS in 3.6% of the patients.
The median volume of the specimens was 229.50 cm 3 with a range of 4 cm 3 to 10 530 cm 3 .

Margin status and management
Of the 192 women who were treated with BCT, 29 had an involved or positive margin after the initial excision (15.1%).
The commonest pathology at the involved margin was carcinoma in situ at (15 patients) 51.7%.Sixteen patients (8.3%) had a close margin of < 1 mm.Sixty-nine patients (35.9%) had a margin between 1 mm and 5 mm.Seventyeight patients (40.6%) had a margin of > 6 mm (Figure 1).Of the patients that had carcinoma in situ at the margins, 8 had repeated WLE, 1 had a mastectomy and 6 had no further surgical intervention.In the infiltrative carcinoma group (14 patients), 6 had a repeated WLE, 2 had mastectomy and 6 had no further surgical intervention (Table 2).
A total of 10 patients had no further surgery for involved margins, 6 and 4 in the DCIS and infiltrative carcinoma groups, respectively.The reasons for no further surgery in the DCIS group were 1 patient developed severe pancreatitis and was lost to follow-up, 4 patients had involved deep margins and 1 patient transferred to another province.In the infiltrative carcinoma group, 1 patient was deemed surgically unfit because of advanced age, 2 patients had tumour at deep margins infiltrating the major pectoralis muscle and we could not determine the reason in the records for 1 patient.
For the women that had a close margin (< 1 mm), 5 and 11 patients had DCIS and infiltrative carcinoma, respectively.In the DCIS group, 3 patients had a subsequent mastectomy and 1 patient each had a repeat WLE and no further surgery because of deep close margins.In the infiltrative close margin group, no patient had a repeat WLE, 2 patients had mastectomy and 9 patients had no further surgery mostly because of deep close margins.
In the group of women who had excision margin 1 mm to 5 mm, one patient (6.3%) had a re-excision for infiltrative    A total of 10 patients had mastectomy for involved or close margins giving an overall mastectomy rate of 5.2% for this study (Table 2).
Residual tumour rate in the re-excision and mastectomy was 50.0% and 63.6%, respectively.The most common residual histology in both groups was DCIS.One patient that had involved margins in the re-excision specimen had a subsequent mastectomy for persistent margin involved.The histology in this patient was IDC in association with DCIS.

Recurrence
Local recurrence is hereby defined as the first recurrence of a tumour in the chest wall or in the operative scar in the ipsilateral breast after the initial WLE.During the median follow-up period of 60 months, a total of 11 patients had histologically confirmed recurrence.A total of 31 patients were lost to follow-up during the observational period, giving the relative observed rate of LR of 6.8%.The median time to recurrence was 39 months.All the cases of recurrences were confirmed by histology.Infiltrative ductal carcinoma was the most common histological type of breast cancer at 63.6% followed by DCIS at 18.2%.
Local recurrence was 6.9% in patients with histologically positive margins, 5.8% in patients with 1 mm-5 mm margins, 6.7% in patients with 6 mm-10 mm margins and 6.1% in patients with > 10 mm margins.No recurrence was noted in the close margin group.There was no statistical difference between the resection margins (Pearson's chi-square 0.890).
The total follow-up duration period was calculated from the date of diagnosis to the discharge date or the date the patient was last seen at the oncology clinic at the time of data collection.
Cross tabulation analysis of the rest of the variables such as age of the patient, tumour pathological size, tumour grade, ER status, presence of lymphovascular invasion and completion of radiotherapy post breast-conserving surgery did not show a correlation with LR (Table 3).The Her-2-neu status was not recorded in 49% of histopathological results and thus not is included in the analysis.

Radiation treatment
A total of 157 patients (81.8%) were referred to the radiation oncology unit for radiotherapy post-surgery, of which 152 (96.8%) completed the radiotherapy course.The total radiotherapy dosages and the patients that received boost radiotherapy dosages were not captured.

Ethical considerations
Approval for the study was obtained from the University of Cape Town's Faculty of Health Sciences Human Research Ethics Committee (HREC REF: 141/2014).

Discussion
This study was conducted to review BCT for breast carcinoma in a single tertiary centre.It is the largest study to date conducted in South Africa.The rate of margin positivity was 15.1%.This rate is comparable to and lower than international studies that vary widely from 3% to 52% 7,8,9,10,11,12,13,14,15,16,17 (Table 4).The wide range in positive margin rates internationally is because of a number of factors.These include inconsistent definitions of a positive margin, differences in the use of intraoperative pathological assessment of margins and variations in the handling of surgical specimens and in pathological sampling of margins.
The type of surgery also affected these numbers, specifically whether surgery was diagnostic or therapeutic in intent.Also, different studies had a wide range of actual number of surgical resections that were performed to generate the published positive margin rates. 18 this study, patients who had tumours with lymphovascular invasion were more likely to have a positive margin (Pearson's chi-square 0.002).Singletary reported that positive margin was significantly associated with large tumour size, young age, axillary node positive status, presence of lymphovascular invasion and an EIC. 19Aziz et al. 20 reported that patients who were younger or had tumours with lymphovascular invasion or a DCIS component were more likely to have a positive margin on a univariate analysis.In this study, the following variables did not predict positive margins: age, pathological tumour size, pathological axillary nodal status, histological diagnosis, grade and ER status.
Nearly, all the tumours in our study are pT1 and pT2 sized (91.7%).This is in keeping with international standards and guidelines for BCT for early breast cancer.However, the current trend internationally is that the eligibility of BCT has been expanded to locally advanced breast cancers post neoadjuvant chemotherapy.Primary tumour response rates of approximately 80% have been observed post neoadjuvant. 21everal RCTs have demonstrated the oncologic safety of neoadjuvant chemotherapy for locally advanced breast cancer.The NSABP B-18 trial demonstrated that there was no statistically significant difference in the LR rate following BCT in the preoperative and postoperative chemotherapy arms. 22is practice is not yet adopted in our surgical oncology unit.Although a total of six patients had WLE post neoadjuvant chemotherapy, we postulate that these patients either had large breast to tumour ratio or their tumours were located high on the chest wall or in the inframammary fold that would have rendered a mastectomy difficult.
Positive resection margins have been associated with a higher LR rate. 5,10,12,19,23It is a standard practice in our unit to re-excise a positive margin with the intent of achieving clear margin prior to radiation therapy.The decision for reexcision is discussed in a multi-disciplinary team comprising surgeons with special interests in breast cancer, medical and radiation oncologists, histopathologists and radiologists.
A variety of factors are taken into consideration before proceeding with re-excision, such as patient's age, comorbidities, life expectancy, extent of excision, extent of margin involvement, tumour characteristics and whether the patient will receive adjuvant treatment.Involved or close margins deep down to fascia are not re-excised because of the morbidity that results from partial muscle excision on the chest wall.Such patients usually receive boost radiation doses.
Our unit's overall re-excision rate of 13.5% is lower than most international studies.An observational study of 2206 women reported an overall re-excision rate of 22.9%.It is notable that within this study, there were wide substantial variations between surgeons and institutions following BCT. 24The most consistent and reliable risk factors for re-excision appear to be the presence of micro-calcifications, EIC or DCIS, and lobular histological type. 25A palpable tumour is more likely to facilitate complete excision, whereas the existence of impalpable [RB.5]DCIS outside the sensible part of the tumour can result in involved resection margin and the subsequent need for a second procedure.
Globally, indications for re-excision for involved or close margins are under constant review.A meta-analysis that included 33 studies and 28 162 patients with ipsilateral recurrences supports the use of no ink on the tumour as an adequate negative margin of resection for invasive breast cancer.The authors concluded that there was no evidence that a wider margin of normal breast tissue than no ink on the tumour decreased the rate of recurrence in the clinical setting of multimodality treatment. 26Re-excision is therefore not mandatory for close margins < 1 mm and each case should be individualised.
In our study, the mastectomy rate post the initial WLE for positive margin is 0.5% and 2.1% for DCIS and invasive cancer, respectively.For close margins, it is 1.6% and 1.0% for DCIS and invasive cancer, respectively.The surgical decision for a re-excision versus a mastectomy is made at a combined breast clinic (CBC).Tumour factors and patient preferences are taken into consideration.Patients with diffuse rather than focally involved margins are a mastectomy.A mastectomy is the preferred surgical treatment for multicentric DCIS or persistent positive margin after repeated WLE.
The residual tumour rate in re-excisions and mastectomy groups is 50% and 63.6%, respectively.These rates are comparable to international studies. 19,27,28In our study, the final margin status measurements were not captured, and therefore, we are not able to report the final close margin rate.However, we noted one patient (0.52%) who had undergone more than one surgical procedure, a re-excision followed by a mastectomy for persistent positive margin.
The median volume of the excised specimen is 229.5 cm 3 in our study.This seems to be larger than the sizes of 60 cm 3 and 70 cm 3 quoted by Vicini et al. 29 and Olivotto et al. 30 The volume of resected breast tissue has a direct impact on cosmesis in breast-conserving surgery.According to Olivotto et al., 30 tumours > 70 cm 3 resulted in a significant increase in the number of cosmetic failures.We postulate that the larger volume reflects the likelihood that our patients presented with slightly larger sized tumours and larger breasts.The cosmetic effects of BCT were not part of this study.A different study in our unit looking into this aspect will provide more insight.
Our study shows a relative observed recurrence of 6.8% at a median follow-up of 60 months.The recurrence rate in the involved or positive margin patient group is 6.8%.Our LR rate is comparable, if not lower, to international studies that show a LR that varies between 2% and 13% for negative margins and 5.5% and 33% for positive margins 7,9,10,11,12,13,14,15,16,17,23 (Table 4).Our data were not sufficient to demonstrate any statistically significant difference between LR and the variables studied.
A total of 156 (81.3%) of the 192 patients that had WLE were referred for radiotherapy.All patients with breast cancer are discussed at the multi-disciplinary team or CBC.At the CBC, the following factors are assessed for BCT for every patient: oncological safety, technical feasibility and delivery of radiotherapy.Contraindications to radiotherapy (previous radiotherapy at this site, connective tissue disorders), if their weight exceeds the radiotherapy weight limit (140 kg) and whether they are physically able to undergo the treatment in the supine position and can attend 15 sessions of RT.Elderly patients over the age of 80 years are not referred for radiation therapy.In our unit, a total of 40.05 Gy to the chest wall is given over 15 hypofractionated schedules (2.67 Gy).A boost (2.67 Gy in 4 schedules) is given to women < 50 years or positive margin in whom repeat surgery is not feasible.There were patients that were lost to follow-up after their surgery because of defaulting, some followed up with private radiation oncologists, some moved to other provinces and some declined further treatment.Of the referred patients, 151 (96.8%) completed radiation treatment.
Our study has a number of limitations.Firstly, this was a retrospective study based on the local hospital sample.The database may not have included all the patients that had WLE and incomplete histological reports.We could not access information on non-retrieved folders.This lack of records may have influenced the final outcomes of the study and it reflects the general challenges in the public health sector with regard to poor record-keeping.Therefore, the analysis is subject to selection bias.The recurrence rate was calculated taking into consideration the patients that are lost to follow-up during the observational period.The use of adjuvant systemic therapy and endocrine therapy is not reported.In general, standard regimens for adjuvant chemotherapy in our unit include adriamycin, cyclophosphamide and paclitaxel in patients < 70 years of age.It would have provided an opportunity to compare the recurrence percentages between arms.Furthermore, we acknowledge that the 18% and 49% of patients without recorded ER and Her-2-neu receptor status is high.This limitation is because of the fact that routine recording of ER and, in particular, human epidermal growth factor receptor 2 (HER-2) receptor status for all patients only became standard practice in our pathology department towards the end of the study period.

Conclusion
Our study reviews the outcomes of BCT women with early breast cancer in a single tertiary centre.It represents the largest cohort of patients managed at a single institution in South Africa.Apart from the inherent limitation to this retrospective review, the median follow-up period of 60 months is reasonable and provides valuable information regarding our unit's surgical margins in BCT, re-excision and ipsilateral breast tumour recurrence rate.The analysis of 192 patients demonstrates outcomes that appear well in keeping with those reported in the international literature with regard to the LR rate.

FIGURE 1 :
FIGURE 1: Bar chart showing the percentages of the excision margins.

TABLE 1 :
Patient and tumour characteristics.

TABLE 2 :
Excision margins and procedure after initial wide local excision.none had a mastectomy.Neither re-excision nor mastectomy was performed in the women with excision margins of > 6 mm.
Source: Author's own work carcinoma and

TABLE 3 :
Effect of different variables on local recurrence.
Source: Author's own work Level of statistical significance is 0.05.It is not significant if > 0.05.ER, oestrogen receptor.

TABLE 4 :
Studies on positive margins and local recurrence by margin status.
a , Recurrence in focal or single positive margin patients.b , Recurrence in multiple positive margins patients.c , Recurrence in patients with positive margins in the current study.d , Overall (i.e.inclusive of positive margins and those that did not have).