Prognostic value of bone marrow biopsy in operable breast cancer patients at the time of initial diagnosis: Results of a 20-year median follow-up.

From May 1975 until May 1980,128 operable breast cancer patients, clinical stage I-II, had a core bone marrow biopsy (BMB) from the posterior iliac crest as a part of the routine diagnostic work-up at the time of initial diagnosis. The mean age of the patients was 56 years, range 26-93. In a previous study on this material, 10 patients (7.8 per cent) were positive for tumor cells and 118 negative by conventional histopathology of BMB [1]. In 1996 we reexamined all BMB separately at two laboratories, using monoclonal antibodies against cytokeratins AE1-AE3, KL1, CAM 5-2 (DOP), and DC10, BA17 (MCI). The number of extrinsic cells in the bone marrow was graded positive for micrometastases when > or = 5 cells or suspicious when 1-4 cells per approximately 2 x 10(6) bone marrow cells were found, using high power field magnification. Micrometastases were detected in 17 patients (13.3 per cent) and another 8 patients were classified as suspicious. The presence of micrometastases was correlated to the axillary lymph node stage and primary tumor location. Median follow-up was 20 years. All 17 micrometastatic patients relapsed and died within 6 years of disease progression with evident osseous metastases. There was one disease-free survivor of the 8 patients with suspicious BMB after 17 years of follow-up. The median overall survival was significantly shorter in tumor-cell positive patients, being 1.9 years compared to 11.7 years in the BMB negative and BMB suspicious groups (p < 0.0001). Immunohistochemical analysis of core BMB taken postoperatively may be useful in predicting the prognosis in patients with breast cancer clinical stage I-II.