Loco-regional recurrence after mastectomy in high-risk breast cancer--risk and prognosis. An analysis of patients from the DBCG 82 b&c randomization trials.

BACKGROUND AND PURPOSE

In the DBCG 82 b&c trials, 3,083 patients with stages II and III breast cancer were randomised to receive post-mastectomy radiotherapy (RT) versus no RT in addition to systemic therapy. The study showed a decrease in loco-regional recurrences and an improved survival in patients receiving RT. The aim of the present study was to identify risk factors for loco-regional recurrence (LRR), to evaluate the treatment of LRR and to examine the prognosis after LRR.

PATIENTS AND METHODS

The 18-year probabilities of LRR were calculated for different prognostic factors using the Kaplan-Meier method. The efficacy of different LRR treatments was compared. The 5-year survival and distant metastases (DM) probability after LRR was calculated with regard to initial randomization group, primary tumor and recurrence related variables.

RESULTS

Of the 3,083 patients, 535 had a LRR alone as first site of failure. In univariate analyses, large primary tumor size, ductal carcinoma, high malignancy grade, fascia invasion, few removed nodes, many positive nodes and extracapsular invasion were all risk factors for developing LRR. Combined treatment with surgery and RT at the time of LRR increased the persistent loco-regional control. The 5-year probability of subsequent DM was 73% irrespective of initial randomization group. In multivariate analysis, large primary tumor size, many positive nodes, extracapsular invasion, supra/infraclaviculary failures, multiple LRR and a short interval less than 2 years to first LRR were poor prognostic factors for survival.

CONCLUSIONS

Twenty-seven percent of LRR patients had no DM 5 years after failure. Initial randomization group did not alter the prognosis after LRR. Combined treatment of the LRR with surgery and RT improved persistent loco-regional control compared with surgery or RT alone.