Postmastectomy Radiation Therapy

Historical Perspective on Postmastectomy Radiation Therapy

PMRT Prior to Systemic Therapy

The use of postmastectomy radiation therapy (PMRT) for improvement of breast cancer–specific outcomes has long been a topic of intense interest within the field of oncology. Prior to the advent of commonly utilizing systemic therapy in the adjuvant setting, an initial meta-analysis of eight randomized trials comparing mastectomy (radical or simple) with or without PMRT published by Cuzick and colleagues in 1987 demonstrated that PMRT was associated with worse overall survival (OS), in large part owing to an excess in cardiac deaths 10 years after PMRT. In a subsequent update published in 1994, the excess of cardiac deaths was offset by a reduction in breast cancer mortality (BCM) with PMRT, suggesting that PMRT is important not only in reducing local-regional recurrence (LRR), and emphasizing the need for improved PMRT techniques aimed at reducing cardiac dose.

PMRT in the Setting of Systemic Therapy

Up until the late 1990s, PMRT was used mostly in patients with a large burden of nodal disease (four or more nodes) and/or those with large tumors (>5 cm), based largely on retrospective data. For example, a pooled analysis was performed by Fowble and colleagues of 627 women with node-positive disease treated on Eastern Cooperative Oncology Group adjuvant chemotherapy trials from 1978 to 1982 after mastectomy but without adjuvant radiation. Larger tumor size (>5 cm), four or more positive lymph nodes, estrogen receptor (ER) negativity, tumor necrosis, and pectoral fascial involvement were found to be significantly associated with LRR within 3 years. Therefore it was postulated that the addition of PMRT to adjuvant chemotherapy would help decrease LRR in patients with certain high-risk features and potentially lead to a demonstrable increase in OS.

In the late 1990s, data from multiple prospective randomized controlled trials (RCTs) emerged that showed a benefit from PMRT in terms of reduction in LRR and improvement in OS in the setting of systemic therapy. Three of these landmark studies included the Danish British Cooperative Group (DBCG) 82b and 82c studies, and the British Columbia (BC) study. The DBCG 82b and 82c studies were conducted from 1982 to 1989 and 1982 to 1990, respectively, and the BC study enrolled patients from 1979 to 1987.

In DBCG 82b, 1708 premenopausal high-risk women (defined as positive axillary lymph nodes, >5 cm primary, and/or invasion of the skin or pectoral fascia) who underwent mastectomy with axillary lymph node dissection (ALND) were randomized to adjuvant chemotherapy with cyclophosphamide-methotrexate-fluorouracil (CMF) for nine cycles versus CMF for eight cycles plus PMRT, which consisted of 50 Gy in 25 fractions or 48 Gy in 22 fractions with inclusion of the axilla (AX), chest wall (CW), supraclavicular (SCV), infraclavicular, and internal mammary nodes (IMNs) using a five-field arrangement. At a median follow-up of 114 months, there was a significant improvement in 10-year locoregional control (LRC; 91% vs. 68%; P < 0.001), disease-free survival (DFS; 48% vs. 34%; P < 0.001), and OS (54% vs. 45%; P < 0.001) in patients treated with CMF + PMRT, compared with CMF alone.

In DBCG 82c, 1375 postmenopausal high-risk women (defined per 82b) under age 70 who underwent mastectomy with ALND were randomized to PMRT (as per 82b) + tamoxifen or tamoxifen alone. Tamoxifen dose was 30 mg/day for 1 year beginning 2 to 4 weeks after surgery for both cohorts. At a median follow-up of 123 months, PMRT in addition to tamoxifen improved 10-year DFS (36% vs. 24%; P < 0.001) and OS (45% vs. 36%; P < 0.03), as well as decreased the risk of LRR as first site of failure (8% vs. 35%; P < 0.001).

The smaller BC trial randomized 318 premenopausal women with node-positive breast cancer after mastectomy and ALND to either CMF alone for 6 to 12 months or CMF + PMRT, administered between cycles 4 and 5. PMRT consisted of 37.5 Gy in 16 fractions to the CW with opposed tangents and IMNs with a direct field along with 35 Gy in 16 fractions to the SCV nodes through anterior-posterior fields. In this study, 20-year LRC (90% vs. 74%, P < 0.002), DFS (48% vs. 30%; P < 0.001), and OS (47% vs. 37%; P < 0.03) were all significantly improved in the PMRT arm.

Older Consensus Guidelines

Based in large part on the DBCG 82b, 82c and BC studies, the 2001 American Society of Clinical Oncology (ASCO) guidelines for PMRT included the recommendation to include this treatment modality for patients with four or more positive lymph nodes with T1–2 tumors, with insufficient evidence to recommend this treatment in those with one to three positive AX nodes and similar tumors. This is despite the fact that the proportion of patients with one to three positive AX nodes ranged from 58% to 60% on the DBCG 82b, 82c, and BC trials. They did, however, recommend PMRT for patients with T3N1 disease. Of note, although these recommendations included the SCL lymph nodes in the target volume, IMNs were not included due to reported insufficient evidence. Again, this is despite the fact that the IMNs were included as part of the target volumes in the DBCG 82b/82c and BC studies.

Postmastectomy Radiation Therapy in the Era of Modern Systemic Therapy: A Potential Systemic Effect of Pmrt

As noted earlier, the ASCO 2001 PMRT guideline left open questions about the role of PMRT in certain subgroups of lower risk patients with T1–2 disease and one to three positive nodes. Several more “modern” studies have provided guidance for these scenarios. As we will see, however, publication of these large studies cannot keep up with the rapid changes in modern practice, and there remain open questions about optimal patient selection for PMRT.

A 2005 publication from the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) reported a benefit from PMRT to local recurrence (LR) and BCM in women with any node-positive disease. It did not, however, account for whether adequate lymph node dissection of levels I and II were completed and whether the number of positive nodes plays a role in the benefit of PMRT. A follow-up EBCTCG meta-analysis looked at patient-level data from 22 randomized clinical trials that accrued between 1964 and 1986, which included information from 8135 patients. This 2014 publication aimed to ascertain the effect of radiation to the CW, SCL, and/or AX fossa and IMNs after upfront mastectomy and adequate axillary surgery, ultimately including 3786 women. Unsurprisingly there was no benefit of PMRT seen in LR, overall recurrence, or BCM in the 700 patients with node-negative disease. All three of these end points were significantly improved in women with AX node–positive disease. Although the outcomes overall were better in patients with lower AX nodal burden, the absolute benefit of PMRT held regardless of whether patients had either one to three, or four or more, nodes involved. For patients with one to three positive nodes, PMRT led to improvements in LRR (3.8% vs. 20.3%; P < 0.00001) and any recurrence (34.2% vs. 45.7%; P = 0.00006) at 10 years, as well as to an improvement in BCM (42.3% vs. 50.2%; P = 0.01) at 20 years. In the 1133 women who received systemic therapy with cytotoxic chemotherapy or tamoxifen, these benefits continued to hold. A major shortcoming of this study is the age of the trials that were included, which were more likely to include outdated surgical techniques and systemic therapy regimens, as well as higher risk patients without screen detected tumors, all of which could overestimate the absolute benefit of radiation therapy. Conversely, older radiation techniques were utilized, which could have increased radiation-related toxicity. Regardless of these limitations, this study served as the impetus for the ASCO-ASTRO-SSO PMRT guideline update discussed later.

Two more modern trials of PMRT/regional nodal irradiation (RNI) after breast-conserving surgery (BCS) were published in the same issue of the New England Journal of Medicine in 2015. Both of these trials included women with early-stage breast cancer with node-positive or high-risk, node-negative disease. High-risk, node-negative disease was defined differently in each trial. The European Organisation for Research and Treatment of Cancer (EORTC) 22922/10925 phase 3 RCT enrolled patients from 1996 to 2004 and included 4004 patients with stage I to III breast cancer with either AX nodal involvement (~55% in each arm) and/or central/medial primary tumor location with negative nodes (~45% in each arm). Of the patients on the trial, 76% underwent BCS, but mastectomy was allowed, and all patients underwent ALND. Systemic therapy details were not obtained, but 99% and 66.3% of patients with and without nodal involvement, respectively, were recorded to have received systemic therapy. The trial compared whole breast (or CW) radiation alone with inclusion of the medial SCV and IMN (first three to five intercostal spaces depending on location of primary tumor). At 10-year follow-up, nodal radiation led to improved DFS (72.1% vs. 69.1%; P = 0.04), distant metastasis–free survival (DMFS; 78.0% vs. 75.0%; P = 0.02), any first recurrence (19.4% vs. 22.9%; P = 0.02), BCM (12.5% vs. 14.4%; P = 0.02), and a trend toward improved OS (82.3% vs. 80.7%; P = 0.06). Toxicity analysis revealed low rates at 10 years with increased pulmonary fibrosis (4.4% vs. 1%; P < 0.001) and a trend toward increased cardiac fibrosis (1.2% vs. 0.6%; P = 0.06). The 15-year update was published in 2020 and demonstrated a persistent benefit in any breast cancer recurrence (24.5% vs. 27.1%; P = 0.024) and BCM (16.0% vs. 19.8%; P = 0.0055), but no significant difference in DFS, DMFS, or OS.

While the Canadian Cancer Trials Group (CCTG) MA.20 trial was conducted in patients who received BCS, it was also attempting to ascertain whether inclusion of the regional nodes in addition to whole breast radiation improved outcomes, and its results were considered in the development of the updated ASCO-ASTRO-SSO PMRT guidelines. This trial accrued from 2000 to 2007 and included 1832 patients with cT1–3 cN1 breast cancer who underwent BCS and sentinel lymph node biopsy (SLNB, ~39%) or ALND and had positive lymph nodes or patients with AX node–negative disease and high-risk features (>5 cm primary tumor or >2 cm primary tumor with fewer than 10 axillary nodes removed AND grade 3/ER–/lymphovascular space invasion [LVSI]). Adjuvant chemotherapy was delivered per institutional practice prior to radiation therapy and adjuvant endocrine therapy was delivered during or after radiation therapy. RNI was to the axillary, SCV, and IMNs (first three intercostal spaces). At 10 years of follow-up, RNI led to improved DFS (82.0% vs. 77.0%; P = 0.01), including improvement in locoregional recurrence–free survival (LRFS; 95.2% vs. 92.2%; P = 0.009), and DMFS (86.3% vs. 82.4%; P = 0.03). No BCM or OS benefit was demonstrated. Toxicity rates were low, but RNI led to increased rates of grade ≥2 pneumonitis (1.2% vs. 0.2%; P = 0.01) and lymphedema (8.4% vs. 4.5%; P = 0.001). It is important to note that unlike the EORTC 22922 in which just under 50% of the population had AX node–negative disease, only ~10% of the population in MA.20 had AX node–negative disease, so conclusions from this trial are not broadly applicable to patients with high-risk, node-negative disease as defined by this study.

When comparing the results of EORTC 22922 and MA.20 with the EBCTCG meta-analysis, there are a couple of points worth emphasizing. First, the patients treated in the late 1990s/early 2000s had substantially lower 10-year risks of LRR without PMRT/RNI (6.8%–9.5%) compared with the patients with one to three positive AX nodes in the EBCTCG meta-analysis (20.3%). Perhaps the most important finding from the 10-year results of the MA.20 and EORTC 22922 studies was that both showed a statistically significant reduction in the risk of distant metastases (DM) of 3% to 4%. This finding suggests that PMRT/RNI has a systemic effect that cannot be overlooked. While this systemic effect was no longer significant with 15-year follow-up of the EORTC 22922 study, there was a significant reduction in the risk of any recurrence (LRR + DM) and BCM with PMRT/RNI, making it difficult to interpret the 15-year results. There was a significant increase in nonbreast cancer–related deaths with PMRT/RNI (30.9% vs. 26.4%), which underscores the importance of modern radiation therapy techniques to minimize the risk of severe toxicity and death related to radiation dose to the heart (and lung).

The updated EBCTCG meta-analysis, along with the initial publications of the MA.20 and EORTC studies, served as an impetus for ASCO, in partnership with ASTRO and SSO, to publish a PMRT guideline update in 2016. Several updates compared to the 2001 guideline include:

1.
For patients with T1–2 tumors and one to three positive nodes who underwent ALND, they unanimously agreed PMRT improves locoregional failure (LRF), any recurrence, and BCM, making consideration of PMRT in this patient population reasonable. They did stipulate that some patients have such low risk of LRF that benefits may not outweigh risks and individualized decision-making is required, taking into account physician and patient risk tolerance and values. This was still a major change compared to the 2001 consensus guideline.
2.
For patients with T1–2 tumors and positive SLNB without ALND, they advocated for PMRT if it would be recommended regardless of what further nodal assessment showed. If further nodal assessment could influence the decision to recommend PMRT, then ALND should be completed. Of note, this guideline was written when the majority of patients were still receiving completion ALND if SLNB revealed positive nodes, which is becoming less common in modern practice as discussed later in this chapter.
3.
In terms of the nodal levels that should be included for PMRT in patients with T1–2 tumors with one to three positive nodes, they recommended inclusion of apical AX, SCV, and IMNs. The recommendation for inclusion of the IMNs was also a major change compared with the prior 2001 guideline.

The Role of Postmastectomy Radiation Therapy After Neoadjuvant Chemotherapy

In contrast to the role of PMRT after upfront surgery, there currently remain no prospective randomized data to help guide the use of PMRT after neoadjuvant chemotherapy (NAC). In 2008 a consensus statement was released regarding the role of locoregional treatments after NAC. In this statement, it was recommended that PMRT should be considered for patients with clinical stage III breast cancer or for patients with pathologically involved lymph nodes after NAC. These guidelines were based in large part from results of a series of retrospective studies from the University of Texas MD Anderson Cancer Center (MDACC) in the early 2000s that set to clarify the role of PMRT in patients treated with NAC. In summary, these studies showed that patients at highest risk of LRR after NAC and mastectomy and who may therefore benefit from PMRT include those with clinical stage III disease (especially cN2–3 involvement) and those patients who have residual pathologically involved nodes (ypN+) at the time of surgery. More recently, disease response to NAC, particularly nodal response in patients presenting with biopsy-proven, AX node–positive disease, appears to be an important biomarker in the NAC patient population. We will review these data further and summarize current recommendations.

Response to Neoadjuvant Chemotherapy: Impact of ypN0 Disease in Patients With Upfront Stage II–III Breast Cancer

The Collaborative Trials in Neoadjuvant Breast Cancer (CTNeoBC) working group performed a pooled analysis of 11,959 patients on various clinical trials and found that patients who achieved a pathologic complete response (pCR) had a>60% relative reduction in the risk of death, which was even more pronounced in patients with triple-negative breast cancer (TNBC; >80%) and patients with ER–/human epidermal growth factor receptor-2 (HER2)+ disease who received anti-HER2 systemic therapy (>90%). These data underscore the prognostic significance of disease response to NAC.

The importance of the prognostic value of ypN0 disease and, more importantly, the downstaging from cN1 to ypN0 disease on influencing rates of LRR is underscored in an important analysis by Kaufmann and colleagues in 2012 that combined data from the NSABP B18 and NSABP B27 clinical trials. On these trials, patients who underwent neoadjuvant systemic therapy (NST) followed by mastectomy (n = 1071) were mandated not to receive RNI. Most of the patients enrolled had early-stage disease: 55% cT1–2N0, 20% cT1–2N1, 16% cT3N0, and 9% cT3N1. Clinically node-negative patients with ypN0 disease had low rates of 10-year LRR, regardless of the presence of residual invasive disease in the breast (6.5% breast pCR vs. 6.3% no breast pCR for patients with cT1–2 tumors; 6.2% breast pCR vs. 11.8% no breast pCR for patients with cT3 tumors). In the cohort of 32 patients who had cN1 disease and had a pCR (breast and no nodes), no LRR events were seen at 10 years. In the larger subset of 121 patients with cN1 disease who achieved ypN0 disease but with residual disease in the breast, 10-year LRR rates were low, regardless of initial tumor size (10.8% for cT1–2 tumors and 9.2% for cT3 tumors). The presence of residual nodal disease resulted in high LRR rates in patients who presented with cN1 disease (22.4% for cT3 and 17% for cT1–2). Table 46.1 summarizes additional retrospective series that demonstrate low rates of LRR in patients who achieve ypN0 status and do not receive PMRT. The data from NSABP B18/B27 and these additional studies support prospective evaluation of omission of PMRT in low-risk (cT1–3 cN1) patients with breast cancer who achieve ypN0 status after NAC.

Table 46.1

Summary of Local-Regional Recurrence Rates in Retrospective Studies Evaluating the Role of Regional Nodal Irradiation After Preoperative Systemic Therapy Based on Nodal Complete Response

Study	No. of Patients	Follow-Up (months)	Clinical Stage	% ypN0	LRR (RNI vs. No RNI)	P value
McGuire et al.	106	62	I–II 33% III 67%	100%	5% vs. 10% (7% vs. 33% in stage III)	0.40 (0.04 for stage III)
Le Scodan et al.	134	91	II 63% III 37%	100%	4% vs. 12%	0.12
Shim et al.	151	57	II 60% III 40%	100%	2% vs. 8%	0.15
Wang et al.	142	72	II 100%	100%	5.5% vs. 9.9%	0.15
Huang et al.	490	72.9	II 53% III 47%	100%	5% vs. 9%	0.07

LRR , Local-regional recurrence; RNI , regional nodal irradiation; ypN0 , no residual nodal disease.

Breast Cancer Subtype

The prior studies included patients treated before the importance of biologic subtype was fully recognized and before the advent of modern anti-HER2 therapy. Newer studies are reporting outcomes by breast cancer subtype treated with modern effective systemic therapy. For example, Mamounas and colleagues presented LRR results of the CTNeoBC by breast cancer subtype in 2041 patients treated with mastectomy, approximately one-third of whom received PMRT. The 5-year LRR for all mastectomy patients was 10.4% and varied by response to NAC (3.3% pCR vs. 8.2% ypT1–3 ypN0 vs. 13.1% ypN+). Compared with patients with ER+/HER2−, grade 1–2 disease, all other breast cancer subtypes had significantly higher risk for LRR, especially patients with ER−/HER2+ disease and TNBC. Whereas residual disease after NAC was strongly associated with LRR in ER−/HER2+ and TNBC, response to NAC did not correlate with LRR in ER+/HER2−, grade 1–2 breast cancer patients (0% pCR vs. 7.3% ypT1–3 ypN0 vs. 5.3% ypN+). Similarly, outcomes by subtype were looked at in 1856 patients who received NAC from 2001 to 2007 in the EORTC 10994 and Breast International Group (BIG) 1-00 studies. Relative to patients with ER+/HER2− breast cancer, patients with TNBC (hazard ratio [HR] = 6.4) and HER2+ disease (HR = 6.3 without trastuzumab and HR = 3.4 with trastuzumab) had significantly higher risk of LRR. Fowble and colleagues also reported LRR by breast cancer subtype in a much smaller cohort of 81 patients with clinical stage I–IIIA breast cancer treated with NST, mastectomy, and no RNI from 2004 to 2013. The 5-year LRR rate was 0% in patients with triple-negative disease (n = 19) and HR−/HER2+ (n = 7) disease and 6% for patients with HR+/HER2− disease, but was 25% in patients with HR+/HER2+ disease (14% in 7 patients with pCR, 31% in 11 patients with ypN0 disease and 33% in 3 patients with ypN+ disease), all of whom received anti-HER2 directed therapy.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here