Molecular and Cellular Pathology
British Journal of Cancer (2004) 91, 1813–1820. doi:10.1038/sj.bjc.6602189 www.bjcancer.com
Published online 26 October 2004
Real-time RT–PCR correlates with immunocytochemistry for the detection of disseminated epithelial cells in bone marrow aspirates of patients with breast cancer
I H Benoy1, H Elst1, I Van der Auwera1, S Van Laere1, P van Dam1, E Van Marck1, S Scharpé2, P B Vermeulen1 and L Y Dirix1
- 1Translational Cancer Research Group Antwerp (Lab Pathology University of Antwerp/University Hospital Antwerp, 2650 Edegem; Oncology Centre, General Hospital Sint-Augustinus, 2610 Wilrijk, Belgium), Belgium
- 2Medical Biochemistry, University of Antwerp, 2610 Wilrijk, Belgium
Correspondence: Dr LY Dirix, Oncology Centre St-Augustinus, Oosterveldlaan 24, 2610 Antwerp (Wilrijk), Belgium. E-mail: luc.dirix@pandora.be
Revised 18 June 2004; Accepted 17 August 2004; Published online 26 October 2004.
Abstract
Real-time reverse transcriptase–polymerase chain reaction (RT–PCR) is a technique with the potential of improving the quantification of disseminated epithelial cells (DEC) in haematological tissues due to its exquisite sensitivity. This sensitivity may lead to false positivity. Immunocytochemistry (ICC) is regarded as the standard methodology to diagnose DEC. In this study, detection with ICC was compared with quantitative real-time RT–PCR for CK-19 and mammaglobin (hMAM) mRNA in bone marrow (BM) of patients with metastatic breast cancer (MBC). Bone marrow was aspirated from 14 control patients and from 29 patients with MBC. Mononuclear cells (MNC) were isolated. Immunostaining was carried out with the Epimet kit. Quantitative PCR was performed on the ABI Prism 7700. The CK-19 and hMAM mRNA quantities were normalised against
-Actin and calculated relative to a calibrator sample (relative gene expression). All controls were negative by ICC and for hMAM expression measured by RT–PCR, whereas the median RGE value for CK-19 was 0.57. For the MBC patients, the median RGE for hMAM was 0 and 10 out of 25 (40%) tested positive. Median RGE for CK-19 was 2.9 and 20 out of 25 (80%) tested positive. With ICC, the median value was 1 stained cell per sample, and 15 out of 24 (62%) samples were positive. A correlation was observed between CK-19 and hMAM expression (r=0.7; P=0.0003), and between hMAM expression and ICC (r=0.6; P=0.003). CK-19 expression and ICC (r=0.9; P<0.0001) showed the strongest correlation. Reverse transcriptase–polymerase chain reaction for CK-19 resulted in a higher number of positive BM samples of patients with MBC than ICC. Since an excellent correlation is observed between ICC and RT–PCR, and RT–PCR is probably more sensitive with the advantage of being less observer dependent and thus also more easy to automate, we consider our quantitative real-time RT–PCR method as validated for the detection of DEC in the bone marrow of breast cancer patients.
Keywords:
RT–PCR, breast cancer, bone marrow, circulating tumour cells, cytokeratin-19, mammaglobin
