Multidisciplinary Management of Locoregional Recurrent Breast Cancer

INTRODUCTION

Locoregional recurrence (LRR) from breast cancer represents a heterogeneous class of disease that has significant variation in its prognosis and preferred treatments. Therefore, the optimal management of LRR disease requires multidisciplinary assessment and integration of input from all disciplines involved in breast cancer management. In this review, we identify prognostic variables associated with the treatment outcome of LRR disease; discuss the roles of surgery, radiation, and systemic treatments; and define reasonable treatment algorithms to aid in the evaluation and management of patients.

Unfortunately, there are few prospective clinical trials to provide high-level evidence regarding the optimal therapeutic strategy for patients with LRR disease. Such trials are difficult to design and execute because of the heterogeneity of prognostic factors and previous treatments of patients. In addition, LRR is becoming much less common with improvements in diagnostic and therapeutic management of primary disease. For example, the risk of development of LRR in patients with early-stage disease treated with breast-conserving surgery, whole-breast radiation, and appropriate systemic treatments is approximately 0.5% per year.¹ This rate is almost half that seen in the first generation of breast-conserving therapy (BCT) trials conducted several decades ago, which averaged LRR rates of approximately 1% per year.² A number of reasons have contributed to these improved outcomes, including better diagnostic imaging and tumor localization, improved pathologic techniques and attention to surgical margins, more optimal localization of radiation to intended targets, and increased use of effective systemic therapies. These therapeutic advances have also significantly decreased the rates of LRR in patients treated with mastectomy. In one study that focused on LRR for patients with 1-3 positive lymph nodes treated with mastectomy, the LRR rate in a more contemporary cohort who did not receive postmastectomy radiation was 2.8% compared with a rate of 9.5% in a similar staged cohort that was treated approximately 15 years earlier.³ In this study, the era in which the patient was treated was the most significant predictor of the risk of LRR for the cohort of patients who did not receive initial radiation, with an adjusted hazard ratio (HR) of 0.35 for the more recent era (P < .001).

Table 1 highlights many of the important prognostic factors for patients with LRR. One of the first considerations for patients with isolated LRR is to determine whether the disease represents a true recurrence versus a new primary, particularly for those with an in-breast recurrence. The following factors suggest a new primary: change in the biomarker profile and/or histopathology of the disease, recurrence is in a different quadrant of the breast compared with the primary tumor, and changes in gene-expression profiles. Investigators from MD Anderson Cancer Center evaluated 397 patients with an in-breast recurrence and classified approximately half as having a new primary and half as having a true recurrence (based on tumor location, histologic subtype, and biomarker profile). The new primary cohort had a better 10-year survival and developed metastatic disease less commonly than patients classified as true recurrence but were more likely to develop contralateral breast carcinoma.⁴ A similar study from Yale University compared 70 patients classified as having a new primary, with 60 patients classified as having a true recurrence. Again, those with a new primary had a better overall 10-year overall survival (75% v 55%; P < .0001) and distant disease-free survival (DFS; 85% v 41%; P < .0001).⁵

TABLE 1.

Prognostic Factors for Patients With LRR Disease

Other prognostic variables that correlate with a favorable outcome of LRR treatment concern the availability of known effective therapies.⁶ For example, patients with resectable disease have a better prognosis than those with unresectable disease, patients who can receive radiation have a better outcome than those with recurrences in previously radiated sites, and patients who have the option of receiving effective systemic therapies have better outcomes than those with disease that has demonstrated primary resistance to front-line chemotherapy, endocrine therapy, and/or targeted therapies.

All patients with LRR should undergo biopsy to establish the diagnosis and evaluate the biomarkers of the recurrent disease. In addition, all patients should undergo staging studies to rule out the presence of distant metastases. Patients who are found to have synchronous distant metastases with the LRR should be treated in an analogous fashion to patients with stage IV breast cancer, with the exception that they may require palliative radiation should the LRR become symptomatic. For patients with isolated LRR, we favor a more aggressive strategy of multimodality therapy that attempts to eradicate all sites of disease for potential cure. Genetic counseling and genetic testing may also be important, depending on age, personal and family history, and biologic subtype. If systemic therapy will be used before locoregional treatment, sites of recurrent disease in the breast/chest wall and regional nodes should be marked/clipped for subsequent surgical and/or radiation localization.

LRR After Breast-Conservation Surgery

As previously noted, in-breast recurrences after breast-conserving surgery with whole-breast radiation therapy is uncommon. For example, in the meta-analysis of 33 eligible studies used to develop the Society of Surgical Oncology and American Society for Radiation Oncology consensus guideline on margin status for stage I-II breast cancer, the in-breast recurrence was only 5.3% at 79.2 months of median follow-up.⁷ Although positive margins were associated with a 2-fold increased risk of recurrence, no ink on tumor was established as the appropriate margin because more widely negative margins did not significantly reduce the risk of local recurrence. This study also reported that systemic therapy reduced the risk of local recurrence and that young age (< 40 years) was associated with a higher risk of local recurrence. Several other groups have reported that biologic subtype also affects the LRR risk, with triple-negative and human epidermal growth factor receptor 2 (HER2)-positive tumors having the highest in-breast recurrence rates.^8-11 However, many of these studies were published before routine use of HER2-targeted therapies in patients with HER2-positive disease. The use of neoadjuvant therapy has been increasing for operable breast cancers and Mittendorf et al¹² reviewed the rates of LRR in patients undergoing upfront surgery versus those receiving neoadjuvant chemotherapy at the MD Anderson Cancer Center. They reported that tumor-related factors (grade, estrogen receptor [ER] status, lymphovascular invasion, and so on) were most predictive of LRR and not the sequencing of chemotherapy. Valachis et al¹³ reported on data from 9 different centers around the world (4,125 patients) receiving neoadjuvant chemotherapy followed by BCT. The 10-year local recurrence rate was 6.5%, with higher rates noted in patients with ER-negative disease, clinically positive nodes at presentation, lack of pathologic complete response in the axilla, and pathologic N2 and N3 disease. A time to recurrence of < 2 years was associated with worse overall outcome.

Traditionally, patients with in-breast recurrence after BCT have been managed with completion mastectomy with or without reconstruction. Historically, autologous tissue reconstruction was favored because the radiated skin and pectoralis muscle often cannot accommodate tissue expansion enough to allow for implant-based reconstruction. Now with the availability of acellular dermal matrices for complete coverage of the implant, some surgeons are offering implant-based reconstruction, although complication rates appear to be increased compared with autologous reconstruction. Many groups have reported favorable outcomes with skin-sparing and even nipple-sparing mastectomy in patients with a history of whole-breast irradiation, although higher rates of skin and nipple necrosis have been reported compared with patients without a prior history of radiation.¹⁴

Over the past 2 decades, there has been increasing interest in performing repeat breast-conserving surgery in patients with a favorable breast-size-to-tumor size ratio. Before considering this approach, it is important to determine whether the patient is a candidate for genetic testing based on age at first primary cancer and biologic subtype. It is also important to consult with radiation oncology to determine whether re-irradiation is feasible. Most single-institution studies have shown that repeat breast-conserving surgery is safe. In a systematic review, Walstra et al¹⁵ found that the addition of radiation was important in reducing the risk of a second recurrence after repeat breast-conserving surgery. Arthur et al¹⁶ reported results from a small phase II trial of repeat breast-conserving surgery and 3-dimensional conformal partial breast irradiation. The 5-year estimate of in-breast rerecurrence was 5.2%. The 5-year cumulative incidence of ipsilateral mastectomy was 10.5%. They found that fibrosis, volume loss, retraction, and pigment change were the most important factors that affected cosmetic outcomes. Although this approach appears to be oncologically safe, patients who already have a poor cosmetic outcome from their initial breast-conserving surgery may not be good candidates for repeat breast conservation because there is likely to be additional reduction in volume and worsening of the appearance of the breast. Figure 1 provides an algorithm for the multidisciplinary management approach in patients with an isolated LRR after BCT.

FIG 1.

An algorithm for multidisciplinary management of patients with locoregional recurrent disease (LRR) after breast-conservation therapy. (*) Breast conservation and re-irradiation of the breast should be used selectively and only for those with favorable tumors (small size, favorable biology, prolonged disease-free interval); consider partial breast radiation if prior whole-breast radiation was previously given. HER2, human epidermal growth receptor 2.

LRR After Mastectomy

The prognosis of patients who develop LRR after an initial mastectomy is unfortunately poor. Even in the setting of isolated LRR disease, numerous published datasets suggest that patients have a high risk of developing subsequent metastatic disease. For example, data from the Danish 82b/c trials, which randomly assigned patients to receive/not receive radiation after mastectomy, demonstrated that 73% of patients with an isolated LRR after mastectomy and systemic treatment developed subsequent distant metastatic disease within 5 years.¹⁷ A second similarly designed postmastectomy radiation trial conducted in British Columbia also found a high risk of subsequent metastases after LRR. In this report, of the 39 patients who developed an LRR, 37 eventually developed metastatic disease.¹⁸

Several publications have provided insights into the prognostic factors associated with outcome for patients with LRR after mastectomy. In the 535 patients who developed an LRR after mastectomy in the Danish 82b and 82c randomized trials, the factors associated with a worse outcome included large initial primary tumor size and a high number of positive lymph nodes, extracapsular extension, recurrence in the infraclavicular or supraclavicular regions, and a disease-free interval of < 2 years.¹⁷ A study of 130 patients treated for an isolated chest wall recurrence after mastectomy at the MD Anderson Cancer Center also found that initial nodal status and time to recurrence were important prognostic factors.¹⁹ However, in contrast to the Danish report, these investigators found that the ability to use radiation to treat the recurrence was an independent predictor of improved outcome for patients with a chest wall recurrence.¹⁹ They then used these 3 features as independent prognostic variables and found that the 19 patients in whom these 3 factors were favorable (initial lymph node–negative disease, a 2-year or longer interval before recurrence, and able to receive salvage radiation) had a 5-year overall survival of 86%. In comparison, the 5-year survival rate for the 89 patients who had 1 or 2 unfavorable features was 48%, and all 22 of the patients who had all 3 unfavorable factors died within 5 years of the recurrence (median survival, 16 months).

Although these data are useful for individualizing a patient’s prognosis, it is also important to recognize the limitations of these datasets. Most of the patients included in these publications did not benefit from the advances that have occurred with modern systemic treatments. For example, the Danish and British Columbia trials referenced previously were conducted during the 1980s and 1990s, and none of the patients received treatment with anthracyclines, taxanes, aromatase inhibitors, or HER2-targeted therapies as components of their initial treatments.

Figure 2 provides an algorithm for the multidisciplinary management approach in patients with an isolated LRR after mastectomy. The initial evaluation should include staging to determine the extent of disease with biopsy of all potential sites of disease (skin, chest wall, regional nodes). Genetic testing should be considered for appropriate patients. The team should also determine, based on the molecular subtype of the disease, the extent of the LRR, and the patient’s previous treatment history, and whether the patient is a candidate for chemotherapy, HER2-targeted therapy, and/or endocrine therapy. Finally, the team should evaluate any previous radiation treatment history to determine whether radiation can be used as a component of salvage treatment.

FIG 2.

An algorithm for multidisciplinary management of patients with locoregional recurrent disease (LRR) after mastectomy. HER2, human epidermal growth receptor 2.

If the patient is a candidate for surgery, chemotherapy, and radiation, our preference is to mark the sites of resectable disease and proceed initially with systemic therapy. This gives the clinicians an opportunity to assess response and to switch therapies if minimal response or progression of disease occurs. Unfortunately, there are no prospective clinical trial data to guide the sequencing of treatments for patients with recurrent disease. Although our preference favors using systemic treatment before surgery, it is equally justifiable to proceed with surgery as the initial step if all sites of LRR can be resected. After surgery and chemotherapy, comprehensive locoregional radiation should be used in patients who did not previously receive radiation therapy as a component of their treatment.

Combined modality treatment, if feasible, achieves the best outcome. A study from Taiwan found that patients who were able to be treated with surgery and radiation had better outcomes than those treated with radiation alone (5-year overall survival rates of 62% v 37%; P = .017).⁶ In addition, they reported that those with more comprehensive radiation fields and use of systemic therapy had better outcomes than patients treated with just a limited radiation field. These data support those of an older study from Washington University, which found that a 10-year rate of locoregional control after radiation for an LRR was 63% in those treated with comprehensive locoregional fields in contrast to 18% for those treated with more isolated fields.²⁰ Finally, investigators from MD Anderson Cancer Center who evaluated 159 patients who were treated with radiation for an isolated LRR after mastectomy found that those without gross disease (whether it was removed surgically or whether a complete response was achieved with neoadjuvant chemotherapy) had an improved 5-year local control rate compared with those with gross disease present at the time of radiation (complete response to surgery, 77%, and complete response to chemotherapy, 83%, v gross disease present at time of salvage radiation, 63%).²¹

Patients presenting with bulky, unresectable disease should be considered for initial chemotherapy or endocrine therapy if active systemic agents are available. If the disease responds favorably, some of these patients may become candidates for surgical resection, which then can be consolidated with comprehensive radiation. The prognosis for patients with unresectable LRR disease that fails to respond to initial systemic treatment is poor, and radiation treatments alone are unlikely to render such patients free of disease. Nevertheless, aggressive locoregional radiation is often used to help stabilize the disease and to avoid the significant adverse consequences of uncontrolled growth of locoregional disease. The dose of radiation depends on the presence or absence of gross disease and whether patients have previously received radiation therapy. For patients who have not had radiation therapy and do not have gross disease, we recommend comprehensive treatment to the chest wall and draining lymphatics to a dose of 50-54 Gy, followed by a boost to the chest wall of 60-66 Gy. Hyperfractionated chest wall irradiation does not seem to provide any benefit over that of conventional therapy given once daily.²² Alternative strategies include using concurrent chemoradiation. For example, a small series of 36 patients explored the use of concurrent radiation treatments with taxane chemotherapy and reported a 2-year rate of local-recurrence–free survival of 68% with acceptable toxicity.²³

For patients who develop recurrence in a previously radiated chest wall, the treatment options are more difficult. There have been several small trials exploring re-irradiation with the addition of local hyperthermia therapy. These data were combined in a systematic review and meta-analysis with 627 patients reported from studies comparing radiation plus hyperthermia versus radiation alone.²⁴ In this cohort, the complete response rate was 60% for hyperthermia with radiation compared with 38% for radiation alone. A similar complete response rate was noted in a separate cohort of 1,483 patients included in single-arm studies using hyperthermia and radiation. Notably, 779 of these 1,483 patients had previous radiation treatment histories. Therefore, for patients with an isolated LRR after previous postmastectomy radiation who have a good performance status, hyperthermia with re-irradiation can be considered. Finally, in a multi-institutional review of repeat radiation, Wahl et al²⁵ reported on 81 patients treated with re-irradiation alone, re-irradiation with hyperthermia, or re-irradiation with concurrent chemotherapy. There was an improved outcome for those without gross disease present at the time of radiation compared with those with gross disease and a trend toward an improved outcome when hyperthermia was added. The normal tissue toxicity reported in this report was acceptably low.

Nodal Management in the Setting of LRR

Patients presenting with isolated nodal recurrence are generally considered for surgical extirpation if the disease is resectable. Like the management of an in-breast recurrence after BCT or chest wall recurrence after mastectomy, patients should undergo staging studies to rule out distant metastatic disease, and neoadjuvant systemic therapy should be considered based on receptor status. Historically, most regional recurrences occurred in surgically inaccessible sites, such as the supraclavicular and/or internal mammary lymph nodes. In such circumstances, radiation therapy should also be considered unless prior radiation treatment reached normal tissue tolerance limits. Since the advent of sentinel lymph node surgery, some patients with isolated LRR may have disease limited to the level I/II axilla. A study from the Netherlands that evaluated 16 breast cancer units identified 54 patients who developed a regional recurrence after an initial negative sentinel lymph node surgery.²⁶ With salvage treatment, a 58% 5-year overall survival was achieved. Patients who had ER-negative disease and those who had limited systemic treatments because of previous adjuvant chemotherapy had poorer outcomes.

Management of the axilla and regional nodes in patients with a local recurrence in the breast or chest wall depends on whether the patient has clinically negative or clinically positive nodes at the time of the recurrence. Ultrasound of the regional nodal basins can be used to identify any abnormal-appearing nodes, and fine-needle aspiration or core biopsy can be used to confirm disease and deploy marker clips for future localization. Le-Petross et al²⁷ identified features on axillary ultrasound associated with nodal disease, including longer short-axis diameter, longer long-axis diameter, increased cortical thickness, and absence of fatty hilum. Repeat sentinel node surgery has been reported by several groups, with greater success rates in patients with an intact breast (compared with mastectomy) and prior sentinel lymph node surgery (compared with axillary lymph node dissection [ALND]). Poodt et al²⁸ published a systematic review of repeat sentinel node surgery in 1,761 patients with an overall success rate of 64.3%. The sentinel node positivity rate was 18.2%, and of those undergoing completion axillary dissection, 16.1% had additional positive nonsentinel nodes. Extra-axillary nodal drainage was more common in patients with a prior ALND, highlighting the role of lymphoscintigraphy as part of the mapping procedure in these patients. In patients where a sentinel node cannot be identified, consideration should be given to complete ALND, especially if identification of nodal disease would affect systemic therapy decisions.

Systemic Treatments for Patients With LRR

Data regarding optimal systemic therapy recommendations and sequencing of systemic treatments and local therapies in the setting of an isolated LRR are limited. For patients with unresectable disease, systemic therapy should be the initial approach. This typically includes patients with supraclavicular and/or internal mammary recurrences and chest wall recurrences that require bony chest wall resection for complete extirpation. Using systemic therapy first can convert some patients to having resectable disease and leave a smaller tumor burden for patients who are candidates for postoperative radiation. In addition, retrospective analyses demonstrate superior outcomes for supraclavicular nodal recurrences treated with chemotherapy plus locoregional therapy.^29,30 Finally, initial systemic therapy in these high-risk situations allows for monitoring of early evidence of distant disease, thus facilitating the identification of patients who can be spared the morbidity associated with locoregional therapies. The exact systemic therapy regimens used should consider prior therapies, medical history, and functional status. Initial endocrine therapy can also be considered for patients with hormone receptor–positive HER2-negative disease if disease burden precludes upfront resection, comorbidities limit use of chemotherapy, and the recurrence has low-risk luminal biology.

For patients who undergo initial resection of an isolated LRR, there are data from a randomized trial that can aid in decision making with respect to systemic therapy. The CALOR trial was an open-label, randomized trial for patients with completely excised isolated LRR.^31,32 One hundred sixty-two patients (58 with ER-negative and 104 with ER-positive recurrences) were randomly assigned to investigator’s choice adjuvant multidrug chemotherapy or no chemotherapy. For hormone-sensitive tumors, chemotherapy was followed by endocrine therapy. If the recurrence occurred while a patient was on endocrine therapy, a change in the endocrine therapy was advised. HER2-directed therapy was optional. Radiation was mandatory for those with positive margins and recommended for those who had not received radiation as part of their initial treatment. The primary endpoint was DFS. Overall survival and breast cancer–free interval were secondary endpoints.

The initial outcome assessments were performed after a median follow-up of 5 years and showed a 12% improvement in DFS with chemotherapy versus no chemotherapy (69% v 57%).³¹ Adjuvant chemotherapy was found to be significantly more effective for women with ER-negative disease (5-year DFS HR, 0.32; 95% CI, 0.14 to 0.73; favoring chemotherapy) with a less certain benefit for patients with ER-positive disease (5-year DFS HR, 0.94; 95% CI, 0.47 to 1.89). The final analysis, performed at a median follow-up of 9 years, confirmed the initial finding: chemotherapy significantly improved 10-year DFS in patients with ER-negative isolated LRR (70% for the chemotherapy subgroup v 34% for the no chemotherapy subgroup; HR, 0.29; 95% CI, 0.13 to 0.67).³² No benefit from chemotherapy could be detected in the ER-positive cohort (10-year DFS, 50%, v 59% in patients treated with or without chemotherapy, respectively; HR, 1.07; 95% CI, 0.57 to 2.00). In the subgroup with ER-negative disease, overall survival at 10 years was 73% with chemotherapy versus 53% without chemotherapy (HR, 0.48; 95% CI, 0.19 to 1.20). Overall survival for those with ER-positive disease was 76% versus 66%, respectively (HR, 0.70; 95% CI, 0.32 to 1.55). In a multivariable analysis, the interaction between ER expression and chemotherapy effect was statistically significant only if the ER status of the isolated LRR was considered, highlighting that the receptor status of the isolated LRR, rather than the primary tumor, should guide systemic recommendations.

In general, these data support using adjuvant chemotherapy for patients who have undergone a resection for an isolated ER-negative LRR. Although not tested in the CALOR trial, use of adjuvant systemic therapy (HER2-targeted therapy ± chemotherapy) should also be strongly considered for patients with HER2-positive recurrences. However, for patients with ER-positive recurrences, the benefit of chemotherapy remains unclear, and treatment recommendations should be individualized and consider patient preferences. It would not be unreasonable to consider chemotherapy for motivated patients with luminal B–type recurrences, those with clear endocrine resistance, or those whose prior treatment lacked a contemporary drug option. Finally, an older randomized trial that investigated tamoxifen use versus no systemic therapy after salvage local therapy for patients with LRR after mastectomy found an improvement in 5-year DFS, from 36% to 59% (P = .007), supporting the use of continued endocrine therapy in the management of patients with ER-positive LRR.³³

Treatment plans for patients who present with LRR after treatment of primary breast cancer can represent some of the most challenging decisions confronting oncology teams, in part because of the diversity of previous treatments and heterogeneity of prognostic factors. Some patients have potentially curable disease, and as such, should receive integrated plans that coordinate surgical, radiologic, and systemic treatments.

AUTHOR CONTRIBUTIONS

Conception and design: All authors

Collection and assembly of data: All authors

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Multidisciplinary Management of Locoregional Recurrent Breast Cancer

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).

Thomas A. Buchholz

Consulting or Advisory Role: Patient Resource, Breakthrough Chronic Care, Abivax, Burjeel Cancer Center, Genentech, Roche Diagnostics

Patents, Royalties, Other Intellectual Property: I am named on a patent held by MD Anderson Cancer Center as a co-inventor on a method of radiation for patients with cancer with magnetically optimized high-energy electron radiation

Sonia Ali

Speakers' Bureau: Amgen, bioTheranostics, Tesaro

Kelly K. Hunt

Consulting or Advisory Role: Armada Health Care, Merck

Research Funding: Endomagnetics (Inst), Lumicel (Inst)

No other potential conflicts of interest were reported.