Background: Four hundred thirty-three patients with recurrent breast cancer were treated by the authors at a large medical breast oncology facility from 1976-1982. The median survival time from first relapse (MSFR) for the 193 patients whose survival experience was not confounded by lead-time bias was 26 months. This MSFR is similar to that of most series published in the 1970s and 1980s and is approximately double that of series published in the 1960s.
Methods: In this series, regression analysis identified disease-free interval (DFI), estrogen receptor (ER) status, and dominant disease site as significant prognostic variables, similar to other published series.
Results: In 113 patients with known ER values, DFI, and dominant metastatic sites, a prognostic spectrum of MSFR patterns was identified among combinations of these three variables. The MSFR ranged from 15 months for poor risk patients with negative ER values, visceral dominant sites, and DFI of less than 24 months, to more than 90 months for good risk patients with positive ER values, soft tissue dominant sites, and DFI of more than 24 months. Although menopausal status alone was not a significant prognostic variable in regression analysis, 66% of premenopausal patients had a constellation of "poor" prognostic variables.
Conclusions: This type of prognostic factor analysis at first relapse could help identify subsets of patients who might be considered for aggressive investigational therapies such as high-dose chemotherapy with autologous bone marrow reconstitution.