Chapter 61 Breast Cancer
Stages I and II
Approximately 70% to 80% of patients with stage I or II invasive breast cancer are technically candidates for breast-conserving therapy (BCT).1,2 Six major randomized trials comparing mastectomy with BCT using modern radiotherapy techniques found no differences in disease-free or overall survival rates between the two approaches3–8 
(see web-only Table 61-1 available on the Expert Consult website), which were confirmed by a meta-analysis including additional studies.9
Many questions remain regarding optimal patient evaluation and selection for BCT, irradiation techniques and doses, factors affecting complications and cosmetic outcome, follow-up, and treatment of locoregional recurrences. Several texts discuss these topics in much greater depth.10–13
Pretreatment Evaluation
Imaging
Mammography should always be performed before treatment, although it does not always accurately delineate the pathologic extent of disease. Magnetic resonance imaging (MRI) has substantial false-positive and false-negative rates.14 It may be most valuable for selected patient subgroups, such as patients with tumors larger than 2 cm, infiltrating lobular histologic characteristics, or positive axillary nodes.15 One small study found a lower risk of local failure in patients undergoing pretreatment MRI,16 but two larger studies with longer follow-up did not.17,18
Postoperative mammograms rarely show residual calcifications in patients with uninvolved microscopic margins.19 In a series from Philadelphia, postoperative mammographic imaging made no difference to the recurrence rate for patients with ductal carcinoma in situ (DCIS).20 Postoperative MRI has limited accuracy in assessing the presence or amount of residual disease.21
Pathologic Evaluation
The most common approach to microscopic margin assessment is surgical removal of a single specimen, which the pathologist rolls in India ink and sequentially sections (as one would slice a loaf of bread) perpendicular to its long axis. Some pathologists “shave” each face of a specimen; the margin is considered involved when tumor is seen in one of the shaved margin slides. These approaches may have quite different clinical implications.22,23 Some surgeons take additional “cavity” specimens after removing the main specimen24; their value is uncertain. The role of intraoperative margin assessment is also uncertain.25
Multidisciplinary Review
One quarter to half of prior community hospital radiologic and pathologic interpretations may be changed on review by specialists.26,27 Management strategies also often differ substantially between physicians and institutions. Hence, it is preferable to have a multidisciplinary review of cases before treatment plans are finalized.
Surgical Techniques
Breast-Conserving Surgery
Technical aspects of breast surgery are discussed in depth elsewhere.10,12,13 Studies differ on whether wider gross tissue resection margins reduce local failure rates.28,29 Patients with large cancers or lesions in areas where resection can cause substantial breast distortion may benefit from “oncoplastic” excision.30,31 Reduction mammoplasty has been used prior to radiation therapy for patients with very large breasts with excellent results.32
Axillary Dissection
The axillary nodes were divided by Berg33 into three anatomic “levels”: level I, lateral and inferior to the border of the pectoralis minor muscle; level II, under the pectoralis minor muscle; and level III (also called the infraclavicular nodes), medial and superior to the border of the pectoralis minor muscle. Randomized trials found no differences in axillary failure or survival rates between “limited” dissection (removal of level I or level I and II nodes only) and “complete” dissection (removal of level I, II, and III nodes), but limited dissection caused less morbidity.34–36
Axillary failures are rare in patients treated with dissection whether nodes are involved or not.33,37 Failure rates are higher in patients treated with inadequate surgery or in patients with eight or more positive nodes.38
Sentinel Node Biopsy
Sentinel node biopsy (SNB) uses injection of a radionuclide tracer or vital dye (or both) in the breast to guide the surgeon in determining which nodes to remove.39 Immediate and long-term complications are substantially lower for SNB compared with level I or II axillary dissection.40–42 False-negative rates of SNB are 0% to 12%, but the risk of axillary failure after negative SNB is very small.43–46 Treatment of patients with involved sentinel nodes is controversial and is discussed below. There were no differences in distant failure rates in two published randomized trials comparing axillary dissection and SNB.47,48 Results from many other trials are pending.
Breast-Conserving Surgery with Radiation Therapy
Patient Status
Pregnancy
Unterminated pregnancy is an absolute contraindication to breast irradiation because of the possible teratogenic and carcinogenic effects on the fetus of scattered radiation.49
Prior Treatment with Radiation Therapy
Although most physicians recommend mastectomy,50 there were no unusual acute or chronic sequelae among 12 women treated with conservative surgery and radiation therapy at the University of Pittsburgh 10 to 29 years after irradiation for Hodgkin’s disease or non-Hodgkin’s lymphoma,51 37 patients treated in Milan,52 or 2 patients treated in Ottawa, Canada.53 However, one of two such patients treated at Stanford University developed severe soft tissue necrosis.54 A few such patients have been successfully treated with partial-breast irradiation, albeit with short follow-up.55,56
Rheumatologic Disorders
The effect of rheumatologic disorders on the risk of developing acute and chronic complications after radiation therapy has been controversial. Some investigators have considered the presence of these disorders a relative or absolute contraindication to the use of radiation therapy.1 However, only three small studies have compared complication rates in patients with rheumatologic diseases with a matched “normal” population.57,58,59 These studies found no clear evidence of increased risk (including the one study focusing on patients treated with BCT58), except for patients with scleroderma. Therefore, the author believes that patients with other rheumatologic disorders can be treated safely but those with scleroderma and its variants should not receive radiation therapy if at all possible.
Breast Size
Women with large breasts have more acute skin reactions and long-term retraction and fibrosis than patients with smaller breasts. Results can be improved by using higher-energy photons,60 1.8-Gy daily fractions, and three-dimensional compensation. Treatment in the lateral decubitus or prone position may also be very helpful (see later discussion).
Prior Breast Augmentation
The risk of capsular fibrosis and other complications following irradiation of patients with prosthetically augmented breasts was 50% or higher in several series61,62 but very low in others.63–65 Local failure rates have also varied between studies, from none65 to 25%.63 The small number of patients in these studies makes it difficult to assess how radiation therapy technique and pretreatment evaluation may affect such outcomes. Given the limitations of mammography in the augmented breast, MRI should be used to help assess the spread of tumor within the breast. If the lesion seems limited after such imaging and margin assessment, it seems reasonable to offer such patients BCT with irradiation. Three-dimensional treatment planning may then help reduce the risk of complications (see section on Acute and Chronic Complications).
Patient Age
Patients younger than 35 to 40 years at diagnosis have higher local recurrence rates than older patients for unclear reasons.66 Despite this, mastectomy does not yield superior distant disease-free or breast cancer-specific survival rates.67,68,69 One study found a 5-year actuarial local failure rate of only 3% among 38 patients with margins wider than 2 mm whose tumor did not contain an EIC.70 In another series, there was no difference in the local failure rate between patients younger or older than age 45 years when margins were wider than 2 mm.71 In a study at the Beth Israel Deaconess Medical Center in Boston, the 8-year local failure rate for 54 patients age 40 years or younger with histologic grade 1 to 2 tumors was 7%, compared with 18% for 74 patients with grade 3 tumors.72 Systemic therapy also reduces local failure rates.66,73
Older patients tolerate radiation therapy well and have excellent local control rates.74,75
Genetic Factors
Many studies found little or no difference in ipsilateral local failure rates between patients with BRCA1 or BRCA2 gene mutations and unaffected patients.76–78,79,80 Other studies have suggested that BRCA1 or BRCA2 mutation carriers have increased late local failure rates, likely resulting from the development of new primary cancers.81–84 Tamoxifen or oophorectomy may reduce rates of both ipsilateral local failure79 and contralateral new primary cancers.79,85 A recent study found equal distant failure rates for 160 BRCA1 and BRCA2 patients treated with BCT and 213 patients treated with mastectomy, with a mean follow-up of 10 years.86 There is also no convincing evidence that contralateral prophylactic mastectomy improves the outcome.80,87,88
Patients with rare genetic syndromes associated with impairment of radiation damage repair, such as ataxia telangiectasia, are at substantial risk of complications from radiation therapy.89 However, patients heterozygous for BRCA1 or BRCA2 mutations do not seem to have increased toxicities.79,90 Complication rates for patients with ATM heterozygosity were not increased in some studies,91,92 but others suggested adverse results for patients with more than one ATM mutation93 or those with particular single mutations.94,95 Patients with polymorphisms of multiple radiation repair genes may also be at increased risk of complications.94,96–99
Some patients with genetic defects of DNA repair may also be at increased risk of radiation carcinogenesis. In-field cancers have been reported to occur within a few years of radiation therapy in several patients with Li-Fraumeni syndrome.100,101 Limited evidence suggests radiation therapy does not increase the risk of contralateral breast cancers in patients with ATM,102,103 BRCA1, or BRCA2 mutations, however.86
Clinical Factors
Means of Detection
Local recurrence rates are similar for patients with palpable and nonpalpable cancers.104 Patients with nipple discharge do not have higher local failure rates than other patients when resection margins are not involved.105
Tumor Size and Location
Tumor size does not influence recurrence rates after BCT.70 It is difficult, however, to achieve both acceptable cosmetic results and negative margins in most patients with tumors larger than 4 to 5 cm without using neoadjuvant chemotherapy (see Chapter 62).
A few studies suggest that patients with medial tumors have higher local recurrence rates,106,107 but this has not been confirmed. Patients with subareolar or periareolar lesions that do not directly extend to the nipple or areola have high local control rates following excision with negative margins without nipple-areolar resection.108,109 Even when nipple-areolar resection must be performed, the appearance and texture of the remaining breast are at least as good as these characteristics following reconstruction procedures, and sensation in the remaining breast mound is preserved. Local control rates also appear to be comparable to those achieved with mastectomy.110
Bilateral Breast Cancers
Patients with bilateral breast cancer (either synchronous or metachronous) can be treated successfully with BCT without an increased risk of complications,111,112 even when the ipsilateral internal mammary nodes (IMNs) were initially irradiated.113
Multiple Ipsilateral Primary Cancers
Few patients have more than one independent ipsilateral breast cancer at presentation. Only 6 of a consecutive series of 200 patients (3%) with palpable masses had two or more separate lesions on mammography in one series.114 Eight of 225 selected patients (4%) in another series had biopsy-proven additional cancers in a separate quadrant detected only on MRI.115 Only 3 of 183 mastectomy specimens (2%) in another study had multiple lesions without demonstrable continuity.116 The majority of apparently independent lesions in one recent study were clonal, indicating spread from a single site of origin.117 Local failure rates in patients with multiple ipsilateral primary tumors having negative margins are similar to those of patients with a single lesion.112,118–122
Pathologic Factors
Margin Status
A “positive,” or “involved,” margin is defined as invasive cancer or DCIS present at an inked surface when a “bread-loafing technique” is used (i.e., sequential sections of the specimen are obtained as one would slice a loaf of bread) or as tumor found anywhere in “shaved” margin specimens. Patients with involved margins generally have higher failure rates than those with uninvolved margins. Involvement by either invasive cancer or DCIS has similar implications.123,124
In a study from the Joint Center for Radiation Therapy (JCRT) in Boston123 with a median follow-up time of 127 months, the crude 8-year local failure rate was 14% for 122 patients with “focal” margin involvement (defined as DCIS and invasive cancer across all examined slides that could be encompassed by three or fewer low-power microscopic fields) and 27% for 66 patients with “extensive” involvement. In a study from Marseille, France, the risk of local failure was 14% (10 of 70) for patients with a single positive margin, compared with 36% (17 of 47) for patients with multiple involved margins with a median follow-up of 72 months.125
Many investigators have arbitrarily distinguished “negative” from “close” but uninvolved margins based on a minimum tumor-free margin width. However, only three published studies with long follow-up times have subdivided results according to sequential intervals71,123,126 (Table 61-1). They do not show consistent patterns between increasing margin width and lower local failure rates.
TABLE 61-1 Tumor-Free Margin Width and Risk of Local Failure Following Breast-Conserving Therapy with Radiation Therapy*
Margin width may still be important for certain patient subgroups, such as young patients. Chemotherapy and tamoxifen substantially reduce failure rates in patients with uninvolved margins as a whole,127,128 but this may be proportionally greater for patients with margins narrower than 1 to 2 mm.123,129
One study found that the volume of disease near uninvolved margins affected the risk of local recurrence.126 The number of excisions required to achieve uninvolved margins was not a risk factor for local failure in most,130,131 but not all,132 series.
Other Histologic Features
An extensive intraductal component (EIC) is defined when two features are simultaneously present for infiltrating ductal carcinomas: (1) intraductal carcinoma that is a prominent portion of the area of the primary mass (in an earlier definition, 25% or more) and (2) intraductal carcinoma clearly extending beyond the infiltrating margin or present in grossly normal adjacent breast tissue.133 Predominantly noninvasive tumors, with only focal areas of invasion, are also included in this category. (Some investigators have used other definitions, most of which probably describe the same entity.) Tumors with an EIC extend more widely through the breast than others.134 The presence of an EIC was a major risk factor for local failure in studies in which microscopic margins were not routinely examined, but its impact is uncertain when margin status has been taken into account.70,123,135
The presence of lymphovascular invasion has been an independent risk factor for local recurrence in most,136–138 though not all,139 studies.
Patients with infiltrating lobular carcinomas or mixed ductal and lobular features have recurrence rates comparable to those of patients with infiltrating ductal carcinomas.140–143 Results for rarer histologic subtypes also seem similar.143–147
Benign proliferative diseases do not appear to affect the risk of local failure.148,149 One study found a substantial risk of residual DCIS or invasive disease on reexcision when atypical ductal hyperplasia was at or near the margins, however.150
Studies conflict on whether finding lobular carcinoma in situ (LCIS) does151–153 or does not154–156 increase local failure rates.
Biologic Markers of Tumor Behavior
Patients with tumors that express the estrogen receptor have modestly lower local recurrence rates than patients with estrogen receptor-negative tumors when systemic therapy is not used. For example, the 10-year local recurrence rate for patients with node-negative, estrogen receptor-positive tumors in the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14 trial was 11%, compared with 15.3% for patients with estrogen receptor-negative tumors in the companion B-13 trial.157 Appropriate systemic therapy reduced this difference substantially (rates of 3.6% and 3.5%, respectively, in these trials).
The impact of c-ERBB-2/neu, or HER2, overexpression on the risk of local recurrence is uncertain.158,159 Several studies found similar local recurrence rates for patients with “triple-negative” cancers and patients with other combinations of estrogen receptor, progesterone receptor, and HER2,160–163 but such patients had modestly higher local failure rates in several recent series.164–166
The implications of gene expression profiles for local therapy are just beginning to be explored. The “wound signature score” discriminated between patients age 53 years or younger who had a high or low risk of recurrence in one study.167 The 10-year local failure rate after BCT was low (4% to 5%) for patients age 50 years or older with node-negative, estrogen receptor-positive tumors, regardless of the score of the 21-gene Oncotype DX test (Genomic Health, Redwood City, California), but for younger patients with a low score, the risk was 12.5%, compared with 27.7% and 26.5% for patients with intermediate- and high-risk scores, respectively (Mamounas E, oral presentation, San Antonio Breast Cancer Symposium, December 2005).168
Breast-Conserving Surgery without Radiation Therapy
Web-only Table 61-2 summarizes the largest randomized trials comparing conservative surgery alone and conservative surgery with radiation therapy for relatively unselected patients.5,169–174 These trials had broad entry criteria and required only very small microscopically tumor-free margins. In the most recently published meta-analysis of the Early Breast Cancer Trialists’ Collaborative Group, which included 7311 patients in these and similar trials, the risk of isolated local recurrence among patients not receiving radiation therapy was 32%, compared with 10.3% in those receiving radiation therapy.9 There was a statistically significant reduction in the 15-year breast cancer-specific mortality rate among patients randomized to irradiation, compared with the control arm (35.9% and 30.5% in the two arms, respectively; hazard ratio, 0.83). The overall mortality rate was also significantly reduced (40.5% and 35.2%, respectively). Of note, the absolute reduction in the breast cancer-specific mortality rate was greater for patients with positive axillary nodes than for those with negative nodes, reflecting the higher risk of local relapse without radiation therapy in the former group (Fig. 61-1).
WEB-ONLY TABLE 61-2 Randomized Trials Comparing Breast-Conserving Surgery Alone with Breast-Conserving Surgery and Radiation Therapy for Relatively Unselected Patients with Early-Stage Invasive Breast Cancer, with Median Follow-up Time Longer Than 5 Years*
Many investigators have tried to identify patient subgroups with an acceptably low risk of local recurrence following conservative surgery alone.175 Age older than 65 years,174 the absence of comedo or lobular features,169 and low histologic grade176 were associated with reduced failure rates in series not employing systemic therapy.
Several randomized trials (Table 61-2) and retrospective or prospective studies186,187 found low local failure rates in selected patients treated with conservative surgery and antihormonal therapy without irradiation. None of the randomized trials showed any differences in rates of distant metastases, breast cancer-specific mortality, or overall survival between the arms. Antihormonal therapy may be needed to obtain acceptably low failure rates even in such populations.178,188,189
TABLE 61-2 Local Failure Rates in Randomized Trials Treating Selected Patients with Breast-Conserving Surgery and Antihormonal Therapy with or without Radiotherapy
There are few data on how specific factors or their combinations affect outcome with conservative surgery with antihormonal therapy. There was no difference in local failure rates between patients ages 50 to 60 years, 60 to 70 years, or older than 70 years with tumors 1 cm or smaller with negative axillary nodes in one study in London.190 In the Ontario–British Columbia trial, at a median follow-up of 94 months the actuarial 8-year local recurrence rate for patients age 60 years or older with positive estrogen receptor or progesterone receptor was 7% for 237 patients with tumors 2 cm or smaller and 5% for 107 patients with tumors 1 cm or smaller treated with tamoxifen alone.181 In the Milan study, patients with infiltrating lobular carcinomas had a slightly higher local failure rate (12%, or 9 of 72 patients) compared with patients with infiltrating ductal histology (8%, or 18 of 231 patients).186 In a retrospective study of patients older than 70 years treated in Nottingham, England, with a limited median follow-up of 37.5 months, the risks of local failure were 33% (4 of 12 patients) for margin widths less than 1 mm, 12% (2 of 17 patients) for margins of 1 to 5 mm, and 2% (1 of 54 patients) for margins wider than 5 mm.191
Even patients with low failure rates may sometimes prefer to reduce their risk of local failure yet further by being irradiated.192,193 The effect of radiation therapy on patients’ quality of life is brief and limited.194 Therefore the decision to omit irradiation should be made by the patient and her physicians together, not unilaterally.
When to Use Nodal Irradiation
Axillary Irradiation
Some patients with clinically uninvolved (cN0) axillary nodes may have a low risk of regional nodal failure without specific axillary treatment181,184,186,195 or with breast irradiation alone.186,196–200 Axillary plus breast irradiation results in failure rates of 1% to 3%,199,201,202,203 very similar to those of axillary dissection. Axillary plus breast irradiation resulted in failure rates that were 1% to 2% lower than those of breast radiation therapy alone in two Italian trials.204,205 Axillary radiation therapy is not as effective for patients with clinically suspicious (cN1) nodes,33 who should undergo a level I or II dissection.
Axillary failure rates after dissection in patients with positive nodes are generally low if a minimum of 6 to 10 nodes are removed, unless there is extensive involvement.38,206 Failure rates may be 7% to 10% or higher in unirradiated patients with eight or more positive nodes or T2 to T3 tumors and four or more positive nodes38
