Breast cancer is the most commonly diagnosed malignancy in women in the United States, accounting for an estimated 176,000 new cases in 1999. Most women present without evidence of distant metastases, and, thus, selection of optimal locoregional treatment is important.
Numerous randomized clinical trials involving several thousand women have demonstrated that breast-conserving therapy with wide excision of the primary tumor and breast irradiation provides long-term survival rates equivalent to those achievable with mastectomy.[1,2] The National Cancer Institute 1995 consensus statement endorsed breast-conserving therapy as appropriate and preferable treatment in most patients with early-stage breast cancer. Nevertheless, most US women with early breast cancer continue to undergo a mastectomy, apparently due to a combination of patient and physician preferences, as well as logistical and access issues.
Among patients who have a mastectomy, the role of postoperative locoregional radiation therapy to the chest wall and draining lymphatics has been very controversial. While numerous studies demonstrate that locoregional radiation improves locoregional control, the data regarding its effects on overall survival are conflicting. Some trials suggest an improvement in overall survival,[5-14] while others do not.[2,15-19]
Some of the conflicting results stem from the fact that certain radiation therapy techniques can increase the risk of cardiac disease.[16,18-22] Thus, some of the radiation-induced reduction in death from breast cancer is negated by an increase in radiation-induced cardiac deaths.[16,22]
Two trials recently updated in The New England Journal of Medicine reported improvements in overall survival with locoregional radiation therapy and have renewed enthusiasm for this approach.[8,11] This article reviews the rationale for locoregional radiation therapy and discusses the results of these two trials, as well their implications for our understanding and treatment of breast cancer.
Mastectomy and axillary dissection (usually levels 1 and 2) does not remove all locoregional disease in all patients. The probability of the surgeon not removing all microscopic cancer may be gauged approximately by the rate of locoregional relapse following mastectomy and is strongly related to the degree of axillary nodal involvement.[8,11,23-28] Recurrences in the chest wall or regional nodal sites (axillary apex, supraclavicular nodes, and, less frequently, internal mammary nodes) occur in approximately 5% to 10% of women with negative axillary nodes but increase steeply with the degree of nodal involvement (Figure 1). The data in Figure 1 are compiled from multiple studies published over several decades that used different methods to segregate patients and calculate the locoregional relapse rate. Despite these limitations, an increasing risk of locoregional recurrence is consistently related to the degree of axillary involvement.
Figure 2 illustrates the results from several randomized clinical trials that documented the locoregional failure rates without (x-axis) and with (y-axis) locoregional radiation therapy.[5,8,11,12,29-33] These data are fairly well approximated by the line y = 1/3 x, suggesting that the addition of radiation therapy reduces the locoregional relapse rate by approximately 67%.
Since the absolute risk of locoregional relapse is highest in patients with multiple positive nodes (Figure 1), it has often been recommended that patients receive locoregional radiation therapy to reduce locoregional recurrence if they have four or more positive axillary nodes.[2,25] This may be a logical approach with respect to locoregional recurrence but not with respect to survival.
If patients have residual regional disease after mastectomy, locoregional radiation therapy can increase the cure rate only if no subclinical distant metastases are present (or if present, if they can be sterilized by systemic chemotherapy). The conventional wisdom of the last few decades has been that breast cancer is a systemic disease, that involvement of regional nodes is a sign of distant spread, and, hence, that treatment of locoregional sites will not affect survival.
These assumptions clearly are not borne out by the data. Node-positive breast cancer is not always a systemic disease. A significant fraction of patients are cured with local therapy alone, without systemic therapy. As summarized in Figure 3, 30-year data from Memorial Sloan-Kettering Cancer Center show that a fraction of patients with small or large cancers and with negative or multiply-positive axillary lymph nodes can be cured without systemic therapy.
Similar data from Milan are shown in Figure 4. Disease-free and overall survival rates at 10 years are ~ 20% to 50% in patients with involved axillary and/or internal mammary lymph nodes, without systemic therapy.
Much of the focus of breast cancer research over the last few decades has been on systemic therapy, leading to a general belief that local treatment is relatively unimportant. This assertion was supported by several studies in early-stage breast cancer showing that lumpectomy plus breast irradiation resulted in a much improved breast control rate than did lumpectomy alone, without marked differences in survival.[1-3,37] Thus, large differences in breast recurrence rates were not readily translated into differences in survival.
However, a recent meta-analysis comparing breast-conserving therapy (lumpectomy plus radiation) with mastectomy suggested that breast conservation (in many respects a more aggressive local therapy than mastectomy) provided a superior survival rate, particularly in node-positive patients. The data suggest that breast cancer is not always a systemic disease, even when regional nodes are involved, and that more aggressive local therapy may improve survival.
Danish Breast Cancer Cooperative Group Trial
Overgaard et al recently published the 10-year update of a trial of 1,473 premenopausal women randomized, following mastectomy, to receive CMF (cyclophosphamide, methotrexate(Drug information on methotrexate), and fluorouracil(Drug information on fluorouracil)) with or without locoregional radiation therapy. All enrolled patients had either positive axillary nodes (90%) and/or a ³ T3 tumor (15%).
The 10-year actuarial survival rate was 54% in the irradiated patients vs 45% in those treated with mastectomy plus CMF alone (P = .001; Figure 5). The magnitude of the survival benefit was 12% in patients with T3, N0 disease, 8% in those with one to three positive axillary nodes, and 12% in those with four or more positive nodes (Table 1).
The irradiation technique included tangential treatment of the ipsilateral chest wall and internal mammary nodes, as well as a separate supraclavicular and axillary field. The radiation dose was 50 Gy in 5 weeks. The extent of the axillary surgery done in this study was less than is typically performed in the United States, with a median number of recovered nodes of seven.
British Columbia Trial
Ragaz et al reported 15-year results of 318 premenopausal women randomized to receive CMF with or without locoregional radiation. The 15-year actuarial survival rate was 54% in the irradiated patients vs 46% in those treated with mastectomy plus CMF (P = .07; Figure 6). Only patients with positive nodes were included in this study. (Unlike the Danish trial, the British Columbia trial excluded patients with T3, N0 disease.)
Precise data on subsets of patients were not provided. However, the authors noted that the magnitude of the relative benefit was similar in the different subgroups. Subgroup analysis relating to the end point of metastasis-free survival is shown in Table 1; the magnitude of benefit was the same in patients with one to three positive axillary nodes as in those with four or more positive nodes.
The extent of axillary dissection done in this study (median of 11 nodes recovered) was comparable to that performed in the United States. The radiation technique was similar to that used in the study conducted by Overgaard et al, although the fractionation scheme was somewhat unusual (37.5 Gy in 3 to 4 weeks).
While the recent presentation of these two studies in The New England Journal of Medicine has renewed enthusiasm for postmastectomy radiation, this really is not totally new information. Early results from the study of
Overgaard et al were published in 1990 in The International Journal of Radiation Oncology Biology and Physics, and the study of Ragaz et al was presented twice at American Society of Clinical Oncology annual meetings in 1993 and 1996.[9,10]