Abstract
The combination of breast conservation surgery (BCS) and whole breast radiation therapy (WBRT) is an effective alternative to mastectomy for the majority of women with early-stage invasive breast cancer, with equal rates of local-regional control and survival. Clinical experience and long-term follow-up of patients have led to improved understanding of clinical and pathologic factors that are important in selecting good candidates for breast conservation therapy. Together with improvements in surgical techniques and the increased use of adjuvant systemic therapy, this greater understanding has helped to minimize the risk of local recurrence. Hypofractionation that reduces length of treatment to 3 to 4 weeks is now an acceptable option for most women requiring WBRT after BCS, although some women based on young age, regional node treatment, or dose homogeneity are suitable for conventional fractionation. Patients who have an ipsilateral breast tumor recurrence (IBTR) can usually be cured by salvage mastectomy, which is the preferred treatment for most women, but there are promising results in studies of further attempts at BCS and additional RT. Although the majority of women will benefit from postoperative WBRT, with a large reduction in the risk of local recurrence, there may be a subgroup of women for whom this benefit is small and clinically acceptable to them. Women of advanced age or substantial comorbidities, who have small tumors excised with wide margins and are treated with endocrine therapy, will have minimal risks of local-recurrence recurrence or breast cancer mortality with the omission of radiation therapy. However, for the majority of women, the use of WBRT after BCS has been associated with a significant improvement in breast cancer mortality and overall survival.
Keywords
breast cancer, breast conservation therapy, radiation, whole breast radiation therapy, hypofractionation
Breast conservation has been a standard alternative to mastectomy for most patients with early-stage 0 to II breast cancer for decades. The purpose of breast conserving surgery (BCS) is the removal of all gross disease, and as much microscopic disease as possible, from the breast while maintaining a good cosmetic result. Residual microscopic disease may then be treated with postoperative radiation therapy (RT), which has been traditionally directed to the entire preserved breast tissue. This combination of BCS and whole breast radiation therapy (WBRT) has been successful in matching the long-term survival as mastectomy. With modern patient selection and multimodality treatment, BCS and WBRT also will result in equally low rates of local recurrence in the preserved breast compared with the chest wall or reconstructed breast after mastectomy.
This chapter examines patient selection, outcomes, and methods of BCS and WBRT for patients with invasive breast cancer and clinically early-stage I and II disease. Additional sections discuss advances in shortening treatment length for RT by hypofractionation, integration of WBRT with systemic therapy, the management of a local recurrence after WBRT, and when BCS alone without RT may be safely offered to selected patients.
Randomized Trials Comparing Breast Conserving Surgery and Radiation Therapy With Mastectomy
Randomized prospective clinical trials have confirmed that BCS and WBRT are associated with equal long-term survival as mastectomy for patients with stage I-II invasive breast cancer ( Table 50.1 ). Meta-analyses of these trials have also confirmed equal distant disease-free survival and overall survival. The equivalence of these two options has been acknowledged by the wide oncology community, since at least 1990 to the present, in published consensus statements by the National Institute of Health, the National Comprehensive Cancer Center Network, and the American College of Radiology. As a measure of quality clinical practice, the National Accreditation Program for Breast Centers of the American College of Surgeons sets a standard of 50% of eligible patients with early stage to be treated with breast conservation.
| OVERALL SURVIVAL (%) | LOCAL RECURRENCE (%) | ||||||
|---|---|---|---|---|---|---|---|
| Trial | Years | No. | Mastectomy | BCS + RT | Mastectomy | BCS + RT | Interval Reported |
| NCI | 1979–87 | 237 | 44 | 38 | 1 | 22 | 25 years |
| Milan | 1973–80 | 701 | 59 | 58 | 2 | 9 | 20 years |
| NSABP | 1976–84 | 1217 | 47 | 46 | 10 a | 14 | 20 years |
| EORTC | 1980–86 | 868 | 45 | 39 | 12 b | 20 b | 20 years |
| Danish | 1983–89 | 793 | 51 | 58 | 21 a | 13 a | 20 years |
| IGR | 1972–79 | 179 | 65 | 73 | 18 | 13 | 15 years |
The risk of a local recurrence in the breast, or ipsilateral breast tumor recurrence (IBTR), in these randomized trials is much higher than generally observed in more modern series. A risk of IBTR of 15% to 20% after BCS and WBRT in a population of stage I and II patients would be unacceptable today. These studies were conducted between the years 1972 and 1987 and do not represent modern patient selection and multimodality (surgery, systemic therapy) treatment. Many factors may account for the high local recurrence in these trials: suboptimal preoperative imaging compared with modern mammography and ultrasound that may have understaged extensive or multicentric breast cancers; less effective surgical-pathologic correlation to imaging findings, particularly for the nonpalpable mass in an era before modern localization procedures; less effective or underutilized systemic therapy compared with modern systemic and targeted therapies; and inclusion of women who may have been at higher risk for new primaries (not distinguishable from local recurrences in most cases) in the preserved breast due to lack of BRCA mutation testing in that era. The surgical techniques and pathologic margin assessment used for BCS in that era also may be largely responsible for the high IBTR rates in those early experiences. A tumorectomy or wide excision was used in most of these randomized trials, except the Milan that used quadrantectomy, but only the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-06 trial required a negative resection margin. The trials with the highest IBTR rates, the National Cancer Institute (NCI) and the European Organisation for Research and Treatment of Cancer (EORTC), permitted patients with tumors of up to 5 cm (compared with upper size limits of 2–4 cm in the other four trials), and allowed gross tumor resection without microscopic margin assessment. For example, in the BCS and WBRT arm of the EORTC trial, 48% of all patients had microscopically positive margins. It is currently recommended that BCS specimens should be oriented for the pathologist to identify the specific margins, and each designated margin evaluated for involvement and closest distance to invasive carcinoma or ductal carcinoma in situ (DCIS).
These randomized prospective trials show that BCS and WBRT that result in comparable rates of local control as mastectomy will yield comparable overall survival for patients with early-stage breast cancer, not that local control is unimportant. The Early Breast Cancer Trialists’ Collaborative Group meta-analysis of randomized trials in the setting of BCS has shown that, as differences in local control rates between two treatments increase, differences in survival may eventually become apparent. Therefore, the goal of BCS and WBRT should be to more closely match or equal the rates of local recurrence of mastectomy—a goal that is now possible with modern multimodality patient selection and treatment.
Patient Selection for Breast Conserving Surgery and Radiation
The rates of IBTR in prospective phase III trials that included in one of the study arms patients treated with BCS and WBRT from a more modern era after 1990 are much lower since the era of the prospective randomized trials comparing to BCS and WBRT to mastectomy from 1972–1989 ( Table 50.2 ). The rates of IBTR of 1% to 3% at 5 years and 2% to 6% at 10 years are now comparable to those achieved by mastectomy for similarly staged patients.
| Trial | Years | N | IBTR (%) | Years at IBTR | Reference |
|---|---|---|---|---|---|
| ACOSOG Z0011 | 1999–2004 | 891 | 2–3 | 5 | |
| PRIME II | 2003–2009 | 658 | 1 | 5 | |
| PMH/BCCA | 1992–2000 | 386 | 4 | 8 | |
| OCOG | 1993–1996 | 1234 | 6 | 10 | |
| CALBG 9343 | 1994–1999 | 317 | 2 | 10 | |
| START A | 1999–2002 | 2236 | 6 | 10 | |
| START B | 1999–2002 | 2215 | 4 | 10 | |
| MA.20 | 2000–2007 | 1832 | 4 | 10 |
The excellent rates of local control with BCS and WBRT are achieved by means of careful patient selection for BCS, pathologic evaluation, radiation, and systemic therapy. Many factors influence the risk of IBTR, and some of the most important clinical and pathologic factors are closely interrelated and cannot be examined in isolation. For example, young patients (35–40 years old or less at diagnosis) have been reported in older series to have a high rate for IBTR. Yet this age group is also more likely than older women to have an extensive intraductal component (EIC) and close or positive resection margins and also more likely to have been treated with undiagnosed BRCA mutation resulting in an apparent higher IBTR rate. In other cases, treatment-related factors, such as extent of surgery or use of systemic therapy, may mitigate the adverse impact of one or more negative prognostic factors. Yet patients whose expected risk of IBTR is very high may still be best served by mastectomy so as not to compromise patient survival. Patient selection for WBRT should also take into account the perceived likelihood of complications from radiation and of an acceptable cosmetic outcome of the preserved breast.
Clinical Factors
There are relatively few clinical patient-related or tumor-related factors that would preclude BCS and WBRT. Young patient age is not alone a contraindication and the risk of IBTR has been markedly reduced in the modern era to low and acceptable levels, comparable in tumor control and survival to mastectomy. Many clinical features or aspects of tumor biology that are associated with an increased risk for IBTR in young women may be more effectively mitigated by optimal patient selection and surgical and systemic therapies. For patients with initial tumor size greater than 5 cm, selected patients with a good response to neoadjuvant chemotherapy may still be candidates for BCS and WBRT. True gross multicentric disease, or diffuse malignant-associated calcifications that cannot be removed, remain indications for mastectomy in most cases. However, multifocal breast cancer within a same quadrant/region of the breast can be treated with BCS if done through a single incision with an acceptable cosmetic result to the patient. Similarly, patients with a subareolar tumor location requiring sacrifice of the nipple-areolar complex, or women with small breast size and a large visible volume defect from surgery, may still prefer breast conservation over mastectomy when the subsequent cosmetic appearance is still acceptable to them despite significant asymmetry. Other contraindications to breast conservation with radiation therapy include a history of collagen vascular disease (scleroderma or active lupus with skin involvement, but not rheumatoid arthritis), prior chest wall irradiation, or pregnancy because of the concern for an increased risk of serious complications or severe late radiation effects in the patient or fetus.
Patient Age
Age younger than 35 to 40 years old has been associated with an increased risk of IBTR after BCS and WBRT compared with older patients. In a meta-analysis of randomized trials comparing BCS with or without WBRT, the risk for any first recurrence (local and distant) by age in node-negative women was 5.9% per year for age younger than 40 years, 2.7% per year for age 40 to 49 years, and 1% to 1.9% per year for 50 years and older. There was also an increased recurrence risk in node-positive women younger than 40 years and a higher breast cancer mortality for women younger than 40 years, compared with older women. However, this does not mean that mastectomy should be the preferred treatment in this age group. The risk for local-regional recurrence (with or without simultaneous distant metastases) is also significantly higher after mastectomy in young women. Mortality is also not reduced in young women with mastectomy compared with BCS and WBRT. For example, a population-based study of 9285 women in Denmark showed that women younger than 35 years treated with BCS and WBRT had a greater risk of IBTR than women ages 45 to 49 years (15% vs. 3%) or women age 50 years or older, but death rates compared with mastectomy groups were the same for each age cohort. In a large population-based study from Canada of 965 women aged 20 to 39, 616 patients had BCS and WBRT and 349 had mastectomy. There were no significant differences in 15-year local-regional recurrence, distant metastases, or survival outcomes between the two groups.
The increased IBTR rates in young women are likely due at least in part to their higher likelihood of having more adverse tumor features than older women. Some of these adverse features are classical pathologic features such as high histologic grade, positive or close resection margins, or a more extensive local tumor burden on reexcisions. This increased risk for IBTR in young women seems to be partly or mostly mitigated when correcting for these adverse prognostic factors by greater surgical attention or modern systemic therapy. For example, a recursive partitioning analysis of more than 900 patients treated with BCS and WBRT found age, margin status, and EIC among the strongest predictors for IBTR, with age the most important. However, when considering women aged 35 years or younger, the risk of IBTR at 10 years was only 3% for patients with negative margins (>2 mm) and EIC-negative tumors compared with 34% when margins were close or positive. Similarly, in a study examining outcome in patients age 40 years or younger in the southern Netherlands, van der Leest and colleagues observed that the risk of IBTR at 10 years was 41% in patients who had an incomplete excision, compared with 16% for those with a complete excision ( p = .005), and 22% for patients not receiving systemic therapy, compared with 11% with systemic therapy ( p = .002).
In more recent studies, the worse prognosis for young women has been attributed to a higher incidence of adverse tumor biology in tumor subtyping and gene expression profiling rather than classical pathologic features. For example, in one study young patients with adverse receptor profiles (luminal B, triple-negative, and HER2 positive) had a higher risk for IBTR that was not observed in young patients with a more favorable luminal A pattern. In a multiinstitutional study of 1434 patients treated with BCS and WBRT, young patient age remained significant with tumor subtyping on multivariate analysis for IBTR but none of the other classical pathologic factors such as grade or margins. However, although statistically significant, the elevated risk for IBTR in the youngest cohort ages 23 to 46 years was 5% compared with less than 2% for older cohorts, a much lower difference and of marginal clinical significance particularly compared with historical studies of age cohorts that did not correct for tumor subtype. In another study of 1058 patients treated with BCS and WBRT, where breast cancer subtype was included in the analysis with other clinical and pathologic traditional risk factors, age less than 40 years was not a significant predictor for IBTR at 10 years (90% vs. 94%, p = .08).
Tumor Size
In standard practice, BCS has been limited to patients with tumors approximately 4 to 5 cm so that a resection to negative margins can be achieved with a good cosmetic result. Patients with tumors larger than 5 cm may be able to have BCS with an acceptable cosmetic result in the setting of a large breast size. However most women with a large tumor size greater than 5 cm or tumors less than 5 cm with a large size to breast ratio are best considered for neoadjuvant chemotherapy for tumor downstaging before BCS. Patients who respond well to neoadjuvant chemotherapy may become better candidates for BCS and WBRT with acceptable cosmetic results.
The NSABP B-18 trial randomly assigned patients to receive either neoadjuvant or adjuvant chemotherapy. BCS and WBRT were performed for 33% of patients presenting with clinical T3 tumors who were assigned to have neoadjuvant chemotherapy, compared with only 9% of those randomized to immediate surgery followed by chemotherapy. IBTR rates for all patients undergoing BCS and WBRT in the neoadjuvant and adjuvant arms were 11% and 8%, respectively, which difference was not statistically significant. However, the IBTR rate was higher (16%) for patients who needed to have chemotherapy to become eligible for BCS and WBRT, compared with those initially felt to be candidates for BCS before chemotherapy (10%). In an updated report of pooled data of 1890 patients treated with BCS + WBRT in the neoadjuvant therapy trials NSABP B-18 and B-27, the 10-year cumulative incidence of IBTR was 8.1%. In the multivariate Cox proportional hazards model, clinical tumor size was not a significant independent predictor of IBTR. Significant factors for IBTR were age, clinical nodal status, and pathologic nodal status/pathologic breast tumor response to neoadjuvant chemotherapy. In a randomized trial performed by the EORTC, 23% of patients originally felt to require mastectomy were treated with BCS after neoadjuvant chemotherapy; but there was no significant difference in local recurrence rates between the neoadjuvant and adjuvant chemotherapy arms. Other studies have also shown favorable results with BCS and WBRT. Fifty-two of 84 women with tumors larger than 5 cm responded to neoadjuvant chemotherapy sufficiently to become 3 cm or smaller and were treated with quadrantectomy and WBRT at the National Cancer Institute in Milan, with a 4% rate of IBTR at 8 years. A study conducted by Chen and colleagues at the MD Anderson Cancer Center reported a 5-year IBTR rate of 8% in 110 patients with initial T3–T4 tumors treated by neoadjuvant chemotherapy, BCS, and WBRT. However, careful clinical and pathologic correlation of tumor stage before and after chemotherapy is needed with this approach. For example, marking the tumor location by a clip before neoadjuvant therapy is recommended to localize the area if there is a good response and ensure negative margins.
Gross Multifocal/Multicentric Disease
The presence of clinically or radiologically detected multifocal disease (defined as having two or more tumors in the same quadrant) or multicentric disease (two or more tumors in separate quadrants) has generally been considered a contraindication to breast conservation therapy. Older studies of such patients have reported elevated IBTR rates of 20% to 40%. However, other reports have had more favorable outcomes in highly selected patients with multiple lesions. Hartsell and coworkers observed a 4% risk of IBTR for patients with two gross pathologic tumors that were EIC negative and resected with negative margins. Cho and colleagues reported no local recurrences after breast conserving therapy in 15 patients, 9 of whom had multiple lesions detected preoperatively and 6 of whom had such disease detected only by the gross pathology examination. Oh and coworkers found a 6% IBTR rate at 5 years among 20 patients presenting with clinically multifocal disease who were treated with neoadjuvant chemotherapy, BCS, and radiation. Okumura and colleagues reported only one local recurrence in 34 patients treated with BCS for macroscopically multiple tumors detected either before (26) or during (8) surgery. Of note, 20 of the 34 tumors were demonstrated to be microscopically contiguous. Yerushalmi and coworkers reported results for 300 patients with multifocal or multicentric stage I or II breast cancer aged 50 to 69, EIC negative, and with small tumor size. These women treated by BCS + WBRT had an IBTR at 10 years of 5.5%, which was similar to that of separate cohorts of unifocal breast cancer treated by breast conservation or patients with one or more tumors treated by mastectomy.
Genetic Factors
A known genetic mutation in BRCA predisposing to breast cancer development is a relative but not absolute contraindication to breast conservation. In a series of 87 women with known BRCA1 or BRCA2 mutations, Robsen and colleagues reported a relatively high risk for IBTR (14% at 10 years and 23% at 15 years). However, Pierce and colleagues reported the results of a multiinstitutional case control study which included 160 known BRCA1-2 mutation carriers and 445 sporadic breast cancer control patients. There was no significant difference in the 15-year risk of IBTR between the BRCA1-2 patients and the controls (24% vs. 17%). Of note, the risk of IBTR (and contralateral breast cancers) was reduced substantially among those BRCA1-2 patients who underwent bilateral oophorectomy or who received tamoxifen, compared with those who did not have such additional treatment. Many of the local “recurrences” in patients with BRCA1-2 mutations may be attributable to the development of new primary tumors, rather than a relapse of the original cancer. Sixty percent of IBTRs in patients with BRCA1 or 2 mutations were located in a different quadrant than the index lesion, compared with 29% for control patients. In another report, Pierce and coworkers compared breast conservation to mastectomy in 655 women with known BRCA1 or 2 mutation. The local recurrence rate at 15 years was 23.5% versus 5.5% ( p < .0001), respectively. However, there were no significant observed differences in the incidence of contralateral breast cancer (>40%) or overall survival. Again, in this series most of the apparent IBTR were considered to be most likely new primaries in the preserved breast. There is other indirect evidence that most IBTR in these patients are actually new primaries. Turner and colleagues found that 8 of 52 patients (15%) with IBTR after breast conserving therapy retrospectively were found to have BRCA1 or BRCA2 mutations. The mean time to IBTR in these women was 8.7 years, compared with 4.3 years for their overall population. Also, the patterns of relapse histology (compared with the first tumor) and location in the breast were different for the BRCA mutation carriers than for other patients, suggesting the majority of these IBTRs were actually new primaries arising later in the preserved breast tissue.
Thus breast conservation therapy is a reasonable option for women at high genetic risk who do not choose to have bilateral mastectomy. Their rates of IBTR, or what may be new primaries in the preserved breast, appear to be reduced to acceptable levels by risk reduction strategies. Closer screening than usual that includes annual magnetic resonance imaging in addition to mammogram is warranted after treatment for the BRCA -associated breast cancer patient who chooses breast conservation.
Race
Race is not an independent factor for IBTR after BCS plus RT after controlling for other prognostic factors in most studies. However, other studies have found a higher risk for IBTR in black women compared with white women. The strong link between race and triple-receptor-negative (estrogen receptor [ER], progesterone receptor [PR], and HER2 negative) tumor subtype makes older retrospective studies that do not account for both in their analyses now outdated. It is known that black women have more common presentation with higher stage breast cancer and lower survival and that this is attributed in large part to a higher incidence of triple-negative breast cancer. Among 704 women all classified as triple-receptor-negative breast cancer subtype, Perez and coworkers found that black women had the same local recurrence rate as nonblack women (3.0% vs. 5.3%, p = .15). Black women in their study did have a higher stage at presentation and higher incidence of regional nodal recurrence.
Pathologic Factors
Margin status and breast cancer tumor subtype approximated by receptor expression are the two main pathologic factors that have a significant impact on the observed rates of IBTR after BCS and WBRT in modern series. Of these, only positive margin status has the potential to preclude BCS in some cases. There is an extensive list of other pathologic features of early-stage breast cancer that have been studied in mostly retrospective analyses for an effect on IBTR, including major ones of histologic grade, nuclear grade, presence of an EIC, lymphovascular invasion, axillary nodal positivity, and extranodal tumor extension. These other factors have minimal to no effect on IBTR once multivariate analyses that include margins and tumor subtype are taken into account, so should not significantly affect patient selection for BCS and WBRT.
Margin Status
A “positive” margin is generally defined as the microscopic presence of tumor cells at an inked edge when radial margin processing, or “breadloafing,” is performed, or the presence of any tumor cells in “shaved” margin specimens. In most series of BCS and WBRT, the risk of IBTR is two to three times greater in the presence of a positive margin compared with a negative margin. A negative margin in many studies can be considered one that is without tumor cells at the ink, or simply not positive. This has been the subject of a major consensus statement (discussed subsequently) that will likely increase its more widespread acceptance in clinical practice. However, other studies have tried to make a distinction between patients with varying degrees of negative margins: a “close” margin (defined variously as cancer cells within distances of 1 mm or 2 mm ) or a “negative” (wider than 2 mm) margin. Retrospective series of BCS and WBRT with margins that are close less than 1 to 2 mm have been associated with much greater heterogeneity in outcomes, with an increased risk of IBTR in some studies but not others. For example, Obedian and coworkers found that patients with close or negative margins both had a 10-year IBTR rate of 2%. In another series, Park and colleagues reported the same crude rate of IBTR (7%) at 8 years in patients with either close (≥1 mm) or negative margins. However, in a prospective trial from the same institution, women with close margins randomized to receive chemotherapy before RT had a crude local failure rate of 32%, compared with 4% in patients receiving RT before chemotherapy. Patients with margins wider than 2 mm have more uniformly across studies had a very low risk of IBTR within 10 years after BCS and WBRT.
The effect of nonnegative margins on IBTR has not been uniform in retrospective studies but associated with heterogeneity based on other clinical and pathologic characteristics. Patient age has been shown in many studies to influence the outcome of involves margins. For example, in a recursive partitioning analysis, Freedman and coworkers showed that close or positive margins were more significant for young patients than for patients older than age 55. Leong and colleagues found that the risk of IBTR in patients with a positive margin was 11% for women age 51 years or older, 21% for women ages 35 to 50 years, and 50% for women age 34 years or younger. Jobsen and colleagues reported that for women 40 years old or younger the 5-year local recurrence rate was 37% when margins were positive and 8.4% when they were negative. Other studies have shown that the number of positive margins and extensive (compared with focal) extent of margin positivity may affect the risk for IBTR several-fold. For example, Park and coworkers reported an 8-year risk of IBTR of 27% in patients with extensively positive margins, compared with 14% for patients with focally positive margins (involvement that could be encompassed within three or fewer low-power microscopic fields); the risk was 13% for patients with uninvolved margins. The IBTR rate with positive margins was reported to be substantially more than two to three times higher than negative margins in cases of invasive lobular carcinoma or an EIC. Systemic chemotherapy has been reported to reduce the impact of a positive margin in some studies but not others. In the same study by Park and colleagues, patients with focally positive margins who received systemic therapy had an IBTR rate of 7% at 8 years, compared with 18% of patients who did not have systemic therapy; however, for patients with extensive margin involvement, systemic therapy had little effect (26% and 29% IBTR rates, respectively). Length of follow-up may be important, with 10-year follow-up needed to observe differences between close and negative margins in some series. For example, Freedman and coworkers found no significant difference in the 5-year cumulative incidence of IBTR for patients with a close or negative margin, but by 10 years a significant difference in IBTR rates was apparent (14% vs. 7%, respectively). Such latency periods of 10 years or more in seeing the rates of IBTR with close margins diverge from those of negative margins has also been shown in other series. Some have practiced a policy of escalating doses for close or positive margins in an attempt to reduce the perceived higher risk for IBTR, although this did not significantly improve local control in the largest prospective randomized trial (discussed later).
This heterogeneity of results regarding the association between margin status and local recurrence may be explained in part by the relatively loose relationship between margins and the extent of residual disease in the breast after surgery. It is this degree of tumor burden that is likely the determining factor for local control. Most of the preceding studies do not report the recurrence risk based on the histology of the tumor, the type of tumor involving the margin, the number of involved resection margins, or the anatomic location of the involved margins. The width of a resection margin can be limited to the surgeon by tumor location and anatomy—for example, a posterior margin at the posterior fascia or deep muscle of the chest wall, or an anterior margin just below but not involving the skin. These situations may not have the same risk for residual disease at the margin as cases where the margin is bordering on breast tissue. Margin location at an anatomic boundary such as the pectoralis muscle is a common reason for not reexcising a margin before radiation. There is a wide variation in the incidence of having residual disease in patients undergoing reexcision (20%–60%), which seems to reflect the variability in these other cofactors with margins. Schnitt and colleagues noted residual carcinoma in 88% of specimens identified to have an EIC, compared with 48% of specimens of patients without an EIC in the primary excision ( p = .002). Mai and coworkers studied serially sectioned lumpectomy specimens and found a relation between having extensive DCIS in the specimen, the incidence of margin involvement, and the risk of more extensive margin involvement. Wazer and coworkers found that the simultaneous presence of an EIC and young age (45 years or younger) were more important than initial margin involvement for predicting the presence of tumor on reexcision. Smitt and colleagues also reported that reexcision findings were related to patient age and the presence of an EIC. Only 11% of patients without an EIC who had close margins had tumor found in the reexcision specimen; this rate was only 10% in patients older than over 65 years. Chism and coworkers found that the risk of having a positive reexcision varied from 15% for patients with T1–T2 tumors with close margins, negative nodes, and no EIC to 83% for those with positive margins, EIC, and positive nodes. Goldstein and colleagues found that increasing amounts of carcinoma near the margin, from the presence of minimal disease to more extensive volume, in addition to the margin width itself, was predictive of IBTR.
Houssami and coworkers conducted a meta-analysis of IBTR after BCS and WBRT, and the relation to final microscopic margin status and the threshold distance for a negative margin. Thirty-three studies met entry criteria for the analysis that included 32,363 patients and margin data on 28,162 patients. The studies were almost all retrospective and treated patients during 1979–2001. The median prevalence of local recurrence was only 5.3% across the 33 studies. The odds ratio in 33 studies for local recurrence was 1.96 for positive or close versus negative margins. In a subgroup of 19 studies with more specific available data on margin width, the odds ratio was 1.74 for close versus negative, and 2.44 for positive versus negative margins ( p < .001). However, additional analyses showed that local recurrence did not significantly vary with margin distance of 1, 2, or 5 mm compared with a margin definition of no tumor on ink. In the adjusted model, the odds of local recurrence were associated with margin status ( p < .001) but not with margin distance ( p = .12), and there was no statistical evidence that the odds of local recurrence decreased as the distance for negative margins increased ( p = .21 for trend).
The Society of Surgical Oncology (SSO) and American Society for Radiation Oncology ASTRO) convened a multidisciplinary panel to develop a consensus statement regarding margins for BCS and WBRT. This panel also included representatives of the College of American Pathologists, American Society of Breast Surgeons, American Society of Clinical Oncology, and a patient advocate. The statement was based in large part on the meta-analysis by Houssami and colleagues that served as the primary evidence base, with additional topic-specific literature reviews conducted by participants for questions not addressed in the meta-analysis. The statement concerns stage I–II invasive breast cancer treated by BCS and WBRT, and did not include cases of pure DCIS, neoadjuvant chemotherapy, or accelerated partial breast irradiation. Major recommendations of the consensus statement were that positive margins, defined as ink on invasive cancer or DCIS, are associated with a twofold increase in IBTR. Negative margins (no ink on tumor) optimize IBTR, but wider margin widths do not significantly lower this risk; the routine practice to obtain wider negative margin widths than ink on tumor is not indicated. It should be noted that this consensus was not a call for an end to all reexcisions in selected clinical cases based on clinical judgment. There was little data available for the meta-analysis to compare margins of no tumor on ink versus 1 mm or greater. However, the consensus opinion was that (1) given the absence of significance to the meta-analysis of margin widths greater than 1 mm, (2) the baseline very low risk for IBTR in the meta-analysis as a whole of approximately 5%, and (3) other improvements in patient selection and systemic therapy since the years of the studies, it is reasonable to assume that any benefit to wider margins not detectable by the power of the meta-analysis would not be clinically significant or greater than 1% to 2%. Other recommendations by the consensus group made based on secondary data from prospective randomized trials and retrospective studies included that there was no indication that a wider margin width than no ink on tumor was routinely needed for young patient age, EIC, lobular histology, whole breast hypofractionation (daily fraction size >2 Gy), or unfavorable biologic subtypes of invasive cancer.
Table 50.3 summarizes the consensus recommendations of several major organizations for the use of reexcision in the setting of nonnegative margins after BCS before WBRT. In summary, it is preferable to have a negative margin (no tumor on ink) before radiation to minimize the subsequent risk of IBTR. However, some patients with focal margin involvement are still reasonable candidates for BCS and WBRT, and others may benefit from a reexcision to obtain a wider margin before WBRT, depending on the presence of other risk factors such as age, histology, the location and extent of these margins, and the use of adjuvant systemic therapy.
| Invasive Breast Cancer | Reference | |
|---|---|---|
| American College of Radiology |
| |
| American Society of Breast Surgeons |
| |
| American Society of Clinical Oncology |
| |
| National Comprehensive Cancer Network |
| |
| Society of Surgical Oncology/American Society for Radiation Oncology |
|
Tumor Subtyping
Gene expression profiling by use of microarrays is a technique that allows the simultaneous study of expression of tens of thousands of genes to looks for patterns that correlate with clinical end points. This technique has been used in patients with early-stage invasive breast cancer, and gene expression in many studies has been predictive for risk of distant metastases and mortality. Many studies have examined the use of gene expression profiles and their prognostic effect for local recurrence with varying degrees of success, and none are currently ready for clinical use in patient selection for BCS and WBRT. Mamounas and coworkers examined the 21-gene recurrence score assay Oncotype DX (Genomic Health, Redwood City, CA) that had been successful in predicting distant recurrence for an association with local recurrence. The recurrence score was available from 390 patients with ER-positive and node-negative invasive breast cancer treated on prospective randomized trials by the NSABP with BCS and WBRT. The local-regional recurrence rates were 6.8%, 10.8%, and 14.6% for low, intermediate, and high recurrence scores, respectively ( p = .043). However, there was a larger variation in IBTR by age with women younger than 50 years (approximately 12%–28% by recurrence score) compared with 50 or older (<5% for all recurrence scores).
Without a validated clinical-use assay to identify these tumors prospectively, these different classes of tumors can be roughly distinguished more easily in clinical practice by expression of currently standard ER, PR, and HER2 receptor testing. Luminal A and B subtypes are associated with positive ER. ER-negative tumors are divided into those positive for HER2, or those of “basal-like” expression associated with low to absent ER, PR, and HER2 receptor expression (or triple-receptor negative). There is an incomplete correlation between basal-like tumors and these receptor expression profiles: from 15% to 45% of basal-like tumors express at least one of these markers, and only 85% of triple-receptor-negative tumors are basal-like by expression arrays. In a multiinstitutional study of breast cancer subtype in 1223 women treated by BCS and WBRT, Hattangadi-Gluth and colleagues reported the 5-year IBTR was 0.2% for luminal A, 1.2% for luminal B, 4.4% for triple negative, and 9% for HER2 positive ( p < .0001) that was significant on multivariate analysis. Arvold and colleagues reported that HER2 positive (and hormone receptor negative) and triple negative subtypes were associated with increased IBTR. Univariate analysis of individual receptor expression HER2/neu overexpression is not associated with an increased risk of IBTR. Patients with ER-negative tumors had an increased risk of IBTR in some but not all studies. Triple-negative receptor pattern was significantly associated with local recurrence in some studies but not others.
In summary, breast cancer subtyping may be associated with differences in IBTR after BCS and WBRT. However, a less favorable tumor subtype, such as triple-negative or HER2 positive, should not influence patient selection for breast conservation. Two studies have specifically compared patients with T1–2N0 triple negative breast cancer subtype treated with BCS and WBRT to mastectomy and found no significant difference in local-regional recurrence, distant metastases, or survival.
Other Pathologic Factors
An EIC is defined as the simultaneous presence of intraductal carcinoma which comprises a prominent portion of the area of the primary mass (generally 25% or more) and of intraductal carcinoma clearly extending beyond the infiltrating margin of the tumor or present in sections of grossly normal adjacent breast tissue. Predominantly noninvasive tumors, with only focal areas of invasion, are also included in this category. The presence of an EIC-positive tumor has been associated with an increased risk of IBTR after breast conserving therapy, although its effect is minimized by increasing the extent of surgery and obtaining negative margins. Patients with positive axillary nodes do not have an increased risk of IBTR compared with patients with negative nodes; indeed, in many earlier studies, their risk of IBTR was generally lower than that of patients with negative nodes, due to the common use of adjuvant systemic therapy for node-positive but not node-negative patients. Extracapsular nodal extension is not associated with an increased risk of IBTR. High histologic grade has also had a variable association with the risk of IBTR after BCS and WBRT, with an increased risk reported by some series but not others. Similarly, lymphovascular invasion has been associated with a relatively increased risk of IBTR in many but not all series. Local recurrence rates are similar for patients with invasive lobular and invasive ductal carcinoma so not a factor in patient selection. Provided margins are negative for invasive disease, the presence of LCIS has also not been associated with an overall increased risk of IBTR.
Treatment Factors
Surgery
The extent of BCS may have an impact on IBTR. In the Milan experience, the risk of IBTR was lower in patients undergoing quadrantectomy than tumorectomy, but the cosmetic results of such wide resections tend to be worse than for patients treated with smaller excisions. Therefore more extensive resection beyond that needed to obtain a negative surgical margin is not necessary.
A reexcision may have therapeutic benefit in converting patients with positive margins to negative margins. Patients achieving a negative margin after reexcision have the same risk of IBTR at 10 years as patients with a negative margin obtained after a single excision. However, a negative reexcision does not have therapeutic value. In the study by Chism and colleagues, the 10-year risk of IBTR with close or positive margins not undergoing reexcision was 5%, the same as for women undergoing a reexcision that contained no residual disease. However, perhaps reflecting the higher tumor burden in these patients, the rate of IBTR was 9% in women with residual disease ( p = .038). Neuschatz and colleagues also found that the risk of IBTR at 12 years was the same for those with no residual tumor on a reexcision as for those patients with initial margin widths of 2 to 5 mm. A reexcision for those patients with initially positive margins, or multiple or extensive close margins, may have more of a diagnostic role to detect patients with a very extensive local tumor burden that suggests mastectomy may be a better option than breast conserving therapy. In patients recommended to have a reexcision before WBRT, a directed reexcision of only the initially involved margins can be employed to reduce volume loss and help to preserve an acceptable cosmetic result.
Radiation Boost
The addition of a radiation therapy “boost,” or an extra dose delivered to the excision site after whole-breast irradiation, has been associated with a decreased risk for IBTR compared with giving whole-breast RT alone in randomized trials. In the EORTC trial, IBTR in the boost and no-boost arms were 12% and 16% at 20 years, respectively. The reduction was statistically significant and proportionally independent of age, but was in magnitude greatest in absolute terms for women age 40 years or younger who had the largest baseline risk of IBTR. However, the EORTC trial did not assess margin status in relation to the in situ component, only invasive breast cancer, and neither trial recorded the exact tumor-free margin width, which may have been quite narrow in some cases. Patients with close or positive margins may also benefit from increased doses given to the tumor bed, which have been reported to decrease the risk of IBTR some series but not others. In the randomized trial by the EORTC of boost versus no boost, for a subgroup of patients available for central pathology review, the risk of IBTR for positive margins was 11.1% compared with 9.4% for negative margins. For patients with microscopic positive margins, a higher boost of 26 Gy was associated with a nonsignificant trend to lower local recurrence compared with 10 Gy (10.8% vs. 17.5%, p > .1).
Use of a concurrent boost is a promising method of maintaining a shorter radiation schedule of whole breast hypofractionation, but also gaining the local control benefit of a lumpectomy cavity boost. Phase III trials studying hypofractionated WBI with a concurrent boost over 3 weeks, including Radiation Therapy Oncology Group 1005 and IMPORT (Intensity Modulation and Partial Organ) High trial in the United Kingdom, if successful may make this approach more widely accepted for use.
Adjuvant Systemic Therapy
The risk of IBTR is lower in patients treated with adjuvant systemic chemotherapy. However, adjuvant systemic therapy may delay the time to develop IBTR but not reduce its long-term risk for patients at high risk of local recurrence, such as those with positive margins. Freedman et al found no significant decrease in the IBTR rate at either 5 or 10 years resulting from the use of systemic therapy in patients with negative margins; for patients with positive margins, the use of systemic therapy was associated with a lower rate of IBTR at 5 years, but it did not significantly decrease the 10-year cumulative incidence of IBTR. Adjuvant tamoxifen has also been associated with a decrease in the rate of IBTR.
Patient Selection Factors for Hypofractionated Whole Breast Irradiation
Hypofractionation is the use of larger than conventional 2-Gy fraction size and generally means a course of radiation delivered in a fewer number of fractions in a shorter period of time than conventional fractionation. There have been four prospective randomized phase III clinical trials of whole breast hypofractionated radiation (H-WBRT) conducted in the United Kingdom and Canada that have long-term 10-year results (see Table 50.2 ). These studies were paramount in establishing H-WBRT as a standard of practice and alternative to conventional fractionation worldwide.
The Ontario Clinical Oncology Group (OCOG) trial randomized patients to 42.5 Gy in 16 fractions over 22 days versus 50 Gy in 25 fractions over 35 days. Eligibility factors included invasive breast cancer 5 cm or less, pathologically node negative, and negative resection margins (defined as no tumor on ink). Exclusion factors included clinical T4 tumors, multicentric disease, and a breast size too large for satisfactory radiation therapy such as a maximum width and the posterior border of tangential fields greater than 25 cm (large chest wall separation). Chemotherapy was permitted before radiation. There was no regional nodal or boost radiation. These two radiation schedules were associated with equivalent 10-year IBTR of 6.2% and 6.7%, respectively. H-WBRT was not inferior to conventional fractionation in terms of breast cancer mortality, death from other causes, or overall survival. The late cosmetic appearance was considered good or excellent in approximately 70% of women in both groups. There were similarly no reported differences in 10-year skin and subcutaneous tissue complications.
The Royal Marsden Hospital and Gloucestershire Oncology Center (RMH/GOC), or Standardization of Breast Radiotherapy (START) pilot trial, was a three-arm trial that used 39 Gy or 42.9 Gy in 13 fractions, compared with 50 Gy in 25 fractions, while maintaining the same 5-week treatment length in all arms. Eligibility was T1–3, N0-1, age less than 75 years, invasive breast cancer, and BCS with a macroscopic negative margin. Chemotherapy was permitted before or during radiation. A breast boost and regional node irradiation were allowed. The IBTR at 10 years was 12.1% for 50 Gy, 14.8% for 39 Gy, and 9.6% for 42.9 Gy, the differences between 39 and 42.9 Gy being statistically significant ( p = .027). The most hypofractionated trial arm had a statistically significant change in breast appearance out to 10 years after radiation. There was a statistically significant observed change in breast appearance up to 10 years after radiation more often in the patients receiving 42.9 Gy compared with 39 Gy and 50 Gy (42.3% vs. 27.4% and 35.4%, respectively), but the number with a marked difference remained relatively low in all three arms (10.1%, 3.4%, and 5.6%, respectively). The differences did not reach significance for 39 Gy compared with 50 Gy. The dose for this arm was modified down for the subsequent START A trial.
From this pilot study, the UK Standardisation of Breast Radiotherapy (START) trials A and B were developed. Trial A compared 50 Gy in 25 fractions, 41.6 Gy in 13 fractions, or 39 Gy in 13 fractions all kept at a duration of 5 weeks (by using every-other-day scheduling in the hypofractionated arms). This was similar to the pilot study described earlier, except for a reduction in dose from 42.9 to 41.6 Gy because of the observed increase in late effects. Trial B compared 50 Gy in 25 fractions over 5 weeks versus 40 Gy in 15 fractions over 3 weeks, a schedule commonly used in the United Kingdom at that time and that was more similar to the Canadian randomized trial. Eligibility included T1–3a, N0-1, invasive breast cancer, and BCS or mastectomy with a margin greater than 1 mm. Chemotherapy, regional node irradiation and boost radiation were all permitted. A sequential boost was permitted on these START trials. There were no differences in 10-year IBTR between the hypofractionation arms and standard fractionation. The local-regional recurrence rates for START A were not significantly different between arms (7.4%, 6.3%, and 8.8%, respectively) or for START B (5.5% and 4.3%, respectively). Rates of breast cancer-specific, disease-free and overall survival were also the same. The late effects of breast appearance, breast edema or hardness, or skin changes were generally equal or better with whole breast hypofractionation, and there were no significant differences in late effects of rib fracture, lung fibrosis, ischemic heart disease or brachial plexopathy.
As a response to the publication of phase III data with 5- to 10-year results, ASTRO convened a task force of experts to make consensus recommendations on the use of H-WBRT in early-stage breast cancer. There was consensus that H-WBRT resulted in equivalent results as conventional fractionation, but for a relatively narrow definition of patients meeting the criteria on the randomized trials and accrued in large numbers on those trials. This was women 50 or older, T1-2, N0, BCS, no systemic chemotherapy, radiation to the breast only, and radiation dose heterogeneity of ±7%. There was no consensus for or against high-grade tumors. There was consensus against H-WBRT for patients not included in the trials, such as DCIS, or underrepresented on the trials such as age less than 50 or needing regional node irradiation. The task force concluded that there were “few data to define the indications for and toxicity of a tumor bed boost” in patients treated with H-WBRT. For patients treated with H-WBRT without a boost as was used on the OCOG trial, 42.5 Gy in 16 fractions was recommended. Lastly, optimization of dose homogeneity was recommended, with a strong recommendation for three-dimensional dose compensation and exclusion of the heart from the primary treatment fields with H-WBRT.
In 2012 ASTRO selected H-WBRT for their first out of five recommendations for the Choosing Wisely campaign, an initiative to foster conversations between patients and physicians about treatments that may be overused, unnecessary or potentially harmful. The recommendation made was “Don’t initiate whole-breast radiation therapy as a part of breast conservation therapy in women age ≥50 with early-stage invasive breast cancer without considering shorter treatment schedules.” H-WBRT was selected for this campaign due to the high prevalence of radiation for early-stage breast cancer in clinical practice, the potential convenience to patients, the reduced cost of care, the strength of the medical evidence and the previously published ASTRO consensus guideline. The text of the recommendation makes note that physicians could discuss with patients that H-WBRT is appropriate for selected women (but does not go into further detail about selection factors). Physicians could also make it known to their patients that there is relatively shorter 10-year follow-up data compared with longer results of 20 years or more for conventional WBI fractionation.
Current patient selection for H-WBRT should be broadly applicable to most women with early-stage breast cancer after BCS. Since the ASTRO consensus, there has been additional large phase III data and other high-quality evidence with long-term. Age less than 50 should not be used as a sole reason to exclude patients from H-WBRT. The OCOG trial reported no significant difference in local control by age 50 and older versus less than 50 and that age less than 50 years had an odds ratio of 1.64 (95% confidence interval [CI] 1.26–21.5, p < .001) compared with 50 and older for having a good or excellent cosmetic result. A meta-analysis of the long-term START pilot, A and B trials also found no significant difference in 10-year local control for ages less than 40, 40 to 49, or 50 and older. The analysis of moderate to marked physician-assessed normal tissue effects on the breast (shrinkage, edema, induration or telangiectasia) was also either not significantly different or actually favored hypofractionation by age in the subgroup analysis. Although there were some initial concerns raised about high-grade and H-WBRT, the randomized trials have shown no difference in the local-regional relapse by grade. Left breast cancer was not specifically excluded from the randomized clinical trials of H-WBRT. The START A and B trials specifically did not show an increase in cardiac events for left-sided patients at 10 years with hypofractionation compared with conventional fractionation. The ASTRO task force recommended for H-WBRT exclusion of the heart from the treatment field using basic three-dimensional treatment planning—a good recommendation for any fractionation. Field in field 3D conformal (forward planning) or intensity modulated radiation therapy (inverse planning), prone position, and deep inspiration breath holding are all standard measures today that may be employed today to reduce cardiac dose. From 11% to 36% of patients in the four major studies of H-WBRT had systemic chemotherapy. A meta-analysis of the long-term RMH/GOC, START A and B trials showed no difference in the local-regional relapse by use of chemotherapy. Subgroup analysis of the OCOG trial showed no difference in local control, or association with a good or excellent cosmetic result, between H-WBRT and conventional fractionation for use of systemic therapy in general (chemo and/or endocrine therapy). The sequential administration of chemotherapy before or after radiation should not be a contraindication to delivering H-WBRT. Because of the possible increased soft tissue and cosmetic effects and lack of data in the setting of H-WBRT, concurrent administration of chemotherapy is not recommended. Given the severe nature of the potential complications, the poor response to treatment of plexopathy, and the potential for very late onset, regional node irradiation is currently recommended only with conventional fractionation.
The OCOG trial specifically excluded patients with a breast “deemed too large to permit satisfactory radiation therapy (i.e., the maximum width of breast tissue > 25 cm).” The UK START A and B trials did not specifically exclude patients for a large breast size, but also placed a limitation on the dose inhomogeneity allowed on the central axis during radiation planning of 5% of the prescribed dose (±5%). The dose inhomogeneity and hot spots need to be restricted even more with H-WBRT than with conventional fractionation to minimize risks for complications due to the steep late effects curve associated with equivalent tissue effects from altered dose fractionation. Keeping the dose homogeneity within 105% to 110% should keep the isoeffective dose for late effects comparable between 2- and 3-Gy fraction sizes. There have been significant advances in radiation planning that have been able to improve dose homogeneity and reduce complications. Today 3D planning by computed tomography simulator and either field in field 3D conformal (forward planning) or intensity modulated radiation therapy (inverse planning) can reduce dose inhomogeneity and reduce complications such as edema or negative cosmetic effects on the breast. In addition to these measures, even a large chest wall separation >25 cm that was used as exclusion in the OCOG trial in the past may be mitigated by differences in patient positionings, such as prone, that can reduce the central axis separation and dose inhomogeneity of large-breasted women.
Related posts:
Anatomy of the Breast, Axilla, Chest Wall, and Related Metastatic Sites
Mesenchymal Neoplasms and Primary Lymphomas of the Breast
Therapeutic Value of Axillary Node Dissection and Selective Management of the Axilla in Small Breast Cancers
Radiation Complications and Their Management
Immunologic Approaches to Breast Cancer Therapy
Delayed Diagnosis of Symptomatic Breast Cancer
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
