International consensus reports recommend postmastectomy radiotherapy only to patients at high risk of a local recurrence (LR).

The present analysis included 1000 out of 3083 high-risk breast cancer patients randomly assigned to postmastectomy radiotherapy in the DBCG82 b&c trials. Tissue microarrays had been constructed and sections stained for estrogen, progesterone and HER2 receptors. Median potential follow-up time was 17 years. Endpoints were LR as isolated first event, breast cancer mortality and overall mortality.

Among patients randomly assigned to not receive radiotherapy, three prognostic subgroups of LR risk were found. "The good" defined by at least four out of five favorable criteria (3 positive nodes, tumor size <2cm, Grade 1 malignancy, estrogen or progesterone receptor positive, HER2 negative), "the Poor" defined by at least two out of three un-favorable criteria (>3 positive nodes, tumor size >5cm, Grade 3 malignancy) and finally "the Intermediate" the group in between. The smallest absolute reduction in 5-year LR probability (11%) after radiotherapy was seen for the good prognosis group. A similar absolute reduction in 15-year breast cancer mortality after radiotherapy (11%) was seen. The largest absolute reduction in 5-year LR probability after radiotherapy was seen for the poor prognosis group (36%). However, this large LR reduction did not translate into any reduction in 15-year breast cancer mortality (0%).

Translation of LR reduction into breast cancer mortality reduction after postmastectomy radiotherapy to high-risk breast cancer patients seems to be heterogeneous, with the largest translation occurring within the good prognosis group.