The current role of post-mastectomy radiotherapy as an adjuvant treatment in stage II breast cancer has been challenged. Although adjuvant radiation was once an established therapy in the management of breast cancer, the routine use of systemic therapy in patients at significant risk for distant (and local) dissemination has caused some to question the need for locoregional radiotherapy. Several issues must be considered when weighing the relative merits of adjuvant locoregional radiotherapy:
- Can systemic therapy improve locoregional control comparable to that achieved with radiotherapy, such that the addition of post-mastectomy radiotherapy offers no further benefit?
- ·Are there subgroups of women at high risk for locoregional failure who would particularly benefit from the use of post-mastectomy radiotherapy?
- ·If post-mastectomy radiotherapy improves locoregional control in the presence of chemotherapy, could this have an impact on overall survival?
To begin to understand the importance of post-mastectomy radiotherapy in the current management of breast cancer, it is necessary to first look at the role this modality has played prior to the use of systemic treatment. Also, when formulating current recommendations for the use of post-mastectomy radiotherapy, it is important to extrapolate the lessons learned from the early trials.
Analyses of locoregional failure have shown the chest wall to be the most common site of failur [1-3]. Halverson et al reported the chest wall to be the site of isolated locoregional recurrence in 60% of patients with a locoregional failure, with an additional 10% of patients having the chest wall as a component of failure, so that the chest wall was involved in 70% of the locoregional failures .
The mastectomy scar is at greatest risk for recurrence . Donegan et al reported the scar or grafted area to be the most frequent site of chest wall involvement, accounting for 42% of such recurrences . In most series, the supraclavicular/infraclavicular nodes are the second most common site of locoregional failure, with involvement as a single site in 10% to 20% patients.
Clinical recurrences occur less frequently in the internal mammary nodes or in the axilla. From 5% to 10% of patients present with an isolated failure at each site in an adequately dissected axilla .
Predictors of Locoregional Failure Following Mastectomy Alone
Additional analyses have identified various factors that are predictive of increased locoregional failure.
Positive Axillary Nodes--The finding of positive axillary nodes has been shown consistently to be the major predictor of chest wall failure [4,6,7]. Haagensen reported in one of the earliest series, factors that predicted a local recurrence following mastectomy . Of 935 patients with either stage I or II disease treated with radical mastectomy and followed for at least 10 years, axillary node involvement was closely related to the risk of chest wall recurrence, with recurrence rates up to 42% in patients with four or more nodes positive.
Donegan et al performed a similar analysis of locally recurrent breast cancer following radical mastectomy . Among the 704 women studied, only 6.5% of patients with histologically uninvolved axillary nodes had a local recurrence at 5 years, as compared with 26% of women with pathologically involved nodes. The percentage of local recurrence was directly correlated with the absolute number of positive nodes: Involvement of 1 to 3 nodes was associated with a 10% to 15% recurrence rate at 5 years; this rate increased to 25% to 38% with 4 to 7 positive nodes and to 43% or higher with 8 or more positive nodes.
In a trial from Stockholm, in which women were randomized to post-mastectomy radiotherapy, histologic lymph node status was an independent predictor of local recurrence. The relative risk for local failure was three times greater among women with positive nodes than among those with negative nodes .
Other factors have also been associated with increased locoregional failure following mastectomy in the absence of systemic therapy. The size of the primary lesion has been correlated with local recurrence in some series [4,8-10]. Donegan and colleagues found a direct association between tumor diameter and local recurrence, with 0% of patients with tumors less than 1.0 cm having a local failure, increasing to 16% for tumors 3.0 to 3.9 cm, 27% for lesions 5.0 to 5.9 cm, and greater than 30% for tumors exceeding 6.0 cm . This study did not correct for the presence of positive axillary nodes, however, in determining this association.
In series reported by Rosenman et al  and Valagussa et al , tumor size was strongly correlated with locoregional recurrence in node-positive patients. Pathologic factors found in the Cancer Research Campaign Trial to independently predict for increased locoregional recurrence included the size of the primary lesion .
Distance of the tumor to the pectoralis fascia may also affect local failure. In one series examining local recurrence as a function of the distance to the deep resection margin, seven of eight patients with local recurrence had a margin of 0.5 cm or less .
Other factors that may also have an impact on chest wall recurrence include lymphatic invasion and tumor grade [10,12].
Quality of life can be severely compromised by a locoregional failure following mastectomy. Subsequent rates of locoregional control with radiotherapy following a recurrence have been disappointing. Reported control rates have ranged between 25% and 76%, depending on tumor bulk and resectability, with an average control rate of 50% [1-3,5]. Thus, even with optimal radiotherapy, 50% of patients experiencing a locoregional failure will die with uncontrolled locoregional disease. This argues strongly in favor of the delivery of prophylactic radiotherapy following mastectomy for maximal local control. In recent series in which radiotherapy has been delivered to high-risk patients in conjunction with chemotherapy, isolated locoregional recurrences have been reduced to 5% to 10% .
It has also been suggested that increased locoregional control influences survival. In every surgical series of stage I or II breast cancer to date, locoregional therapy as sole therapy has been curative in a majority of node- negative patients and a minority of node-positive women [4,6,9]. Therefore, a finite percentage of early-stage breast cancers are confined locoregionally and are cured by definitive locoregional treatment. Conversely, a minority of node-negative women and a majority of node-positive women whose disease is locally controlled develop sites of distant dissemination.
Sequential National Surgical Adjuvant Breast Project (NSABP) studies comparing locoregional therapies have shown little correlation between local control and overall survival [14,15]. Based on these results, Fisher et al hypothesized that breast cancer is a systemic disease at presentation, and that local control has little to no impact on survival. An alternative hypothesis proposed by Hellman suggests that breast cancer is a heterogeneous disease in which subsets of women present with systemic disease, whereas others have disease restricted to locoregional sites at diagnosis . In the latter group, locoregional control would surely influence outcome. Therefore, for reasons including both quality of life and, potentially, overall survival, the maintenance of locoregional control is an important goal.