P ostmastectomy radiation therapy (RT) continues to be controver- sial, but perhaps less so than in the past. [1][2][3][4] Studies convincingly demonstrate that RT improves locoregional control (LRC) and disease free survival in lymph node positive patients, [5][6][7][8][9][10][11][12] and perhaps in lymph node negative patients as well. 9,11,13 Breast carcinoma specific mortality is also improved with RT. 1,4 However, an improvement in overall survival (OS) has been more difficult to demonstrate. 2 In a 1995 overview, a nonstatistically significant improvement in OS was reported with the addition of RT. 2 The inability to convincingly demonstrate an improvement in OS with RT is likely due to excess cardiac mortality after RT that counteracts the reduction in breast carcinoma deaths after RT. 1,14 The incidence of such RT-induced mortality is highly dependent on the radiation technique used 10,14 -16 and should be extremely low with careful modern treatment planning. 3,16 -18 Because RT has a clear impact on LRC but a debatable impact on OS, many breast oncologists have settled on a ''compromise'' policy wherein only patients at high risk of locoregional (LR) recurrence are offered RT. Because the LR failure rate increases with the number of positive axillary lymph nodes, 19 RT has been advocated in patients with four or more, but not one to three, positive axillary lymph nodes. This is a rational approach as it pertains to LRC, but might not be appropriate if the goal of RT is to improve OS. Improvements in LRC and OS may not necessarily parallel each other.
The data from six randomized clinical trials can be interpreted to demonstrate an improvement in OS with RT (Table 1). These studies enrolled pre-and postmenopausal women, and some used systemic chemotherapy. In general, only patients with a moderate-to-high risk of LR relapse were included (either T3 -4 primary tumors [American Joint Committee on Cancer, AJCC] or positive axillary lymph nodes). The Stockholm I and Oslo trials also included lymph node negative women, but only the data from the lymph node positive patients are