N. Harbeck et al., TUMOR-CELL DETECTION IN THE BONE-MARROW OF BREAST-CANCER PATIENTS AT PRIMARY THERAPY - RESULTS OF A 3-YEAR MEDIAN FOLLOW-UP, British Journal of Cancer, 69(3), 1994, pp. 566-571
We examined bone marrow aspirates from 100 metastasis-free primary bre
ast cancer patients. In 38/100 patients (38%), tumour cells were detec
ted in the marrow using an immunocytochemical technique with a cocktai
l of two monoclonal antibodies: anti-EMA and anti-cytokeratin. Median
follow-up was 34 months: 15/38 (39%) tumour cell-positive patients hav
e since relapsed, but only 9/62 (15%) tumour cell-negative patients. T
he median interval between tumour cell detection and relapse was 11.4
months. No statistically significant correlation existed between tumou
r cell presence and 'established' prognostic factors. However, relapse
-free survival was significantly shorter in tumour cell-positive patie
nts. Multivariate analysis showed tumour cell presence as a strong, si
gnificant prognostic factor for relapse-free as well as overall surviv
al. We conclude that screening for tumour cells in bone marrow of prim
ary breast cancer patients identifies high-risk patients for early rel
apse. In particular, patients with node-negative tumours who have tumo
ur cells in their bone marrow may require subsequent systemic therapy.